RELEASE.txt 18 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214215216217218219220221222223224225226227228229230231232233234235236237238239240241242243244245246247248249250251252253254255256257258259260261262263264265266267268269270271272273274275276277278279280281282283284285286287288289290291292293294295296297298299300301302303304305306307308309310311312313314315316317318319320321322323324325326327328329330331332333334335336337338339340341342343344345346347348349350351352353354355356357358359360361362363364365366367368369370371372373374375376377378379380381382383384385386387388389390391392393394395396397398399400401402403404405406407408409410411412413414415416417418419420421422423424425426427428429430431432433434435436437438439440441442443444445446447448449450451452453454455456457458459460461462463464465466467468469470471472473474475
  1. HDF5 version 1.13.1 released on 2022-03-02
  2. ================================================================================
  3. INTRODUCTION
  4. ============
  5. This document describes the differences between this release and the previous
  6. HDF5 release. It contains information on the platforms tested and known
  7. problems in this release. For more details check the HISTORY*.txt files in the
  8. HDF5 source.
  9. Note that documentation in the links below will be updated at the time of each
  10. final release.
  11. Links to HDF5 documentation can be found on The HDF5 web page:
  12. https://portal.hdfgroup.org/display/HDF5/HDF5
  13. The official HDF5 releases can be obtained from:
  14. https://www.hdfgroup.org/downloads/hdf5/
  15. Changes from Release to Release and New Features in the HDF5-1.13.x release series
  16. can be found at:
  17. https://portal.hdfgroup.org/display/HDF5/HDF5+Application+Developer%27s+Guide
  18. If you have any questions or comments, please send them to the HDF Help Desk:
  19. help@hdfgroup.org
  20. CONTENTS
  21. ========
  22. - New Features
  23. - Support for new platforms and languages
  24. - Bug Fixes since HDF5-1.13.0
  25. - Platforms Tested
  26. - Known Problems
  27. - CMake vs. Autotools installations
  28. New Features
  29. ============
  30. Configuration:
  31. -------------
  32. - CPack will now generate RPM/DEB packages.
  33. Enabled the RPM and DEB CPack generators on linux. In addition to
  34. generating STGZ and TGZ packages, CPack will try to package the
  35. library for RPM and DEB packages. This is the initial attempt and
  36. may change as issues are resolved.
  37. (ADB - 2022/01/27)
  38. - Added new option to the h5cc scripts produced by CMake.
  39. Add -showconfig option to h5cc scripts to cat the
  40. libhdf5.settings file to the standard output.
  41. (ADB - 2022/01/25)
  42. - CMake will now run the PowerShell script tests in test/ by default
  43. on Windows.
  44. The test directory includes several shell script tests that previously
  45. were not run by CMake on Windows. These are now run by default.
  46. If TEST_SHELL_SCRIPTS is ON and PWSH is found, the PowerShell scripts
  47. will execute. Similar to the bash scripts on unix platforms.
  48. (ADB - 2021/11/23)
  49. Library:
  50. --------
  51. - Add a new public function, H5ESget_requests()
  52. This function allows the user to retrieve request pointers from an event
  53. set. It is intended for use primarily by VOL plugin developers.
  54. (NAF - 2022/01/11)
  55. Parallel Library:
  56. -----------------
  57. - Several improvements to parallel compression feature, including:
  58. * Improved support for collective I/O (for both writes and reads)
  59. * Significant reduction of memory usage for the feature as a whole
  60. * Reduction of copying of application data buffers passed to H5Dwrite
  61. * Addition of support for incremental file space allocation for filtered
  62. datasets created in parallel. Incremental file space allocation is the
  63. default for these types of datasets (early file space allocation is
  64. also still supported), while early file space allocation is still the
  65. default (and only supported at allocation time) for unfiltered datasets
  66. created in parallel. Incremental file space allocation should help with
  67. parallel HDF5 applications that wish to use fill values on filtered
  68. datasets, but would typically avoid doing so since dataset creation in
  69. parallel would often take an excessive amount of time. Since these
  70. datasets previously used early file space allocation, HDF5 would
  71. allocate space for and write fill values to every chunk in the dataset
  72. at creation time, leading to noticeable overhead. Instead, with
  73. incremental file space allocation, allocation of file space for chunks
  74. and writing of fill values to those chunks will be delayed until each
  75. individual chunk is initially written to.
  76. * Addition of support for HDF5's "don't filter partial edge chunks" flag
  77. (https://portal.hdfgroup.org/display/HDF5/H5P_SET_CHUNK_OPTS)
  78. * Addition of proper support for HDF5 fill values with the feature
  79. * Addition of 'H5_HAVE_PARALLEL_FILTERED_WRITES' macro to H5pubconf.h
  80. so HDF5 applications can determine at compile-time whether the feature
  81. is available
  82. * Addition of simple examples (ph5_filtered_writes.c and
  83. ph5_filtered_writes_no_sel.c) under examples directory to demonstrate
  84. usage of the feature
  85. * Improved coverage of regression testing for the feature
  86. (JTH - 2022/2/23)
  87. Fortran Library:
  88. ----------------
  89. -
  90. C++ Library:
  91. ------------
  92. -
  93. Java Library:
  94. -------------
  95. -
  96. Tools:
  97. ------
  98. -
  99. High-Level APIs:
  100. ----------------
  101. -
  102. C Packet Table API:
  103. -------------------
  104. -
  105. Internal header file:
  106. ---------------------
  107. -
  108. Documentation:
  109. --------------
  110. -
  111. Support for new platforms, languages and compilers
  112. ==================================================
  113. -
  114. Bug Fixes since HDF5-1.13.0 release
  115. ===================================
  116. Library
  117. -------
  118. - Fixed a metadata cache bug when resizing a pinned/protected cache entry
  119. When resizing a pinned/protected cache entry, the metadata
  120. cache code previously would wait until after resizing the
  121. entry to attempt to log the newly-dirtied entry. This
  122. caused H5C_resize_entry to mark the entry as dirty and made
  123. H5AC_resize_entry think that it didn't need to add the
  124. newly-dirtied entry to the dirty entries skiplist.
  125. Thus, a subsequent H5AC__log_moved_entry would think it
  126. needed to allocate a new entry for insertion into the dirty
  127. entry skip list, since the entry didGn't exist on that list.
  128. This caused an assertion failure, as the code to allocate a
  129. new entry assumes that the entry is not dirty.
  130. (JRM - 2022/02/28)
  131. - Issue #1436 identified a problem with the H5_VERS_RELEASE check in the
  132. H5check_version function.
  133. Investigating the original fix, #812, we discovered some inconsistencies
  134. with a new block added to check H5_VERS_RELEASE for incompatibilities.
  135. This new block was not using the new warning text dealing with the
  136. H5_VERS_RELEASE check and would cause the warning to be duplicated.
  137. By removing the H5_VERS_RELEASE argument in the first check for
  138. H5_VERS_MAJOR and H5_VERS_MINOR, the second check would only check
  139. the H5_VERS_RELEASE for incompatible release versions. This adheres
  140. to the statement that except for the develop branch, all release versions
  141. in a major.minor maintenance branch should be compatible. The prerequisite
  142. is that an application will not use any APIs not present in all release versions.
  143. (ADB - 2022/02/24, #1438)
  144. - Unified handling of collective metadata reads to correctly fix old bugs
  145. Due to MPI-related issues occurring in HDF5 from mismanagement of the
  146. status of collective metadata reads, they were forced to be disabled
  147. during chunked dataset raw data I/O in the HDF5 1.10.5 release. This
  148. wouldn't generally have affected application performance because HDF5
  149. already disables collective metadata reads during chunk lookup, since
  150. it is generally unlikely that the same chunks will be read by all MPI
  151. ranks in the I/O operation. However, this was only a partial solution
  152. that wasn't granular enough.
  153. This change now unifies the handling of the file-global flag and the
  154. API context-level flag for collective metadata reads in order to
  155. simplify querying of the true status of collective metadata reads. Thus,
  156. collective metadata reads are once again enabled for chunked dataset
  157. raw data I/O, but manually controlled at places where some processing
  158. occurs on MPI rank 0 only and would cause issues when collective
  159. metadata reads are enabled.
  160. (JTH - 2021/11/16, HDFFV-10501/HDFFV-10562)
  161. - Fixed several potential MPI deadlocks in library failure conditions
  162. In the parallel library, there were several places where MPI rank 0
  163. could end up skipping past collective MPI operations when some failure
  164. occurs in rank 0-specific processing. This would lead to deadlocks
  165. where rank 0 completes an operation while other ranks wait in the
  166. collective operation. These places have been rewritten to have rank 0
  167. push an error and try to cleanup after the failure, then continue to
  168. participate in the collective operation to the best of its ability.
  169. (JTH - 2021/11/09)
  170. Java Library
  171. ------------
  172. -
  173. Configuration
  174. -------------
  175. -
  176. Tools
  177. -----
  178. -
  179. Performance
  180. -------------
  181. -
  182. Fortran API
  183. -----------
  184. -
  185. High-Level Library
  186. ------------------
  187. -
  188. Fortran High-Level APIs
  189. -----------------------
  190. -
  191. Documentation
  192. -------------
  193. -
  194. F90 APIs
  195. --------
  196. -
  197. C++ APIs
  198. --------
  199. -
  200. Platforms Tested
  201. ===================
  202. Linux 5.13.14-200.fc34 GNU gcc (GCC) 11.2.1 2021078 (Red Hat 11.2.1-1)
  203. #1 SMP x86_64 GNU/Linux GNU Fortran (GCC) 11.2.1 2021078 (Red Hat 11.2.1-1)
  204. Fedora34 clang version 12.0.1 (Fedora 12.0.1-1.fc34)
  205. (cmake and autotools)
  206. Linux 5.11.0-34-generic GNU gcc (GCC) 9.3.0-17ubuntu1
  207. #36-Ubuntu SMP x86_64 GNU/Linux GNU Fortran (GCC) 9.3.0-17ubuntu1
  208. Ubuntu 20.04 Ubuntu clang version 10.0.0-4
  209. (cmake and autotools)
  210. Linux 5.8.0-63-generic GNU gcc (GCC) 10.3.0-1ubuntu1
  211. #71-Ubuntu SMP x86_64 GNU/Linux GNU Fortran (GCC) 10.3.0-1ubuntu1
  212. Ubuntu20.10 Ubuntu clang version 11.0.0-2
  213. (cmake and autotools)
  214. Linux 5.3.18-22-default GNU gcc (SUSE Linux) 7.5.0
  215. #1 SMP x86_64 GNU/Linux GNU Fortran (SUSE Linux) 7.5.0
  216. SUSE15sp2 clang version 7.0.1 (tags/RELEASE_701/final 349238)
  217. (cmake and autotools)
  218. Linux-4.14.0-115.21.2 spectrum-mpi/rolling-release
  219. #1 SMP ppc64le GNU/Linux clang 8.0.1, 11.0.1
  220. (lassen) GCC 7.3.1
  221. XL 16.1.1.2
  222. (cmake)
  223. Linux-3.10.0-1160.49.1 openmpi-intel/4.1
  224. #1 SMP x86_64 GNU/Linux Intel(R) Version 18.0.5, 19.1.2
  225. (chama) (cmake)
  226. Linux-4.12.14-150.75-default cray-mpich/7.7.10
  227. #1 SMP x86_64 GNU/Linux GCC 7.3.0, 8.2.0
  228. (cori) Intel (R) Version 19.0.3.199
  229. (cmake)
  230. Linux-4.12.14-197.86-default cray-mpich/7.7.6
  231. # 1SMP x86_64 GNU/Linux GCC 7.3.0, 9.3.0, 10.2.0
  232. (mutrino) Intel (R) Version 17.0.4, 18.0.5, 19.1.3
  233. (cmake)
  234. Linux 3.10.0-1160.36.2.el7.ppc64 gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-39)
  235. #1 SMP ppc64be GNU/Linux g++ (GCC) 4.8.5 20150623 (Red Hat 4.8.5-39)
  236. Power8 (echidna) GNU Fortran (GCC) 4.8.5 20150623 (Red Hat 4.8.5-39)
  237. Linux 3.10.0-1160.24.1.el7 GNU C (gcc), Fortran (gfortran), C++ (g++)
  238. #1 SMP x86_64 GNU/Linux compilers:
  239. Centos7 Version 4.8.5 20150623 (Red Hat 4.8.5-4)
  240. (jelly/kituo/moohan) Version 4.9.3, Version 5.3.0, Version 6.3.0,
  241. Version 7.2.0, Version 8.3.0, Version 9.1.0
  242. Intel(R) C (icc), C++ (icpc), Fortran (icc)
  243. compilers:
  244. Version 17.0.0.098 Build 20160721
  245. GNU C (gcc) and C++ (g++) 4.8.5 compilers
  246. with NAG Fortran Compiler Release 6.1(Tozai)
  247. Intel(R) C (icc) and C++ (icpc) 17.0.0.098 compilers
  248. with NAG Fortran Compiler Release 6.1(Tozai)
  249. MPICH 3.1.4 compiled with GCC 4.9.3
  250. MPICH 3.3 compiled with GCC 7.2.0
  251. OpenMPI 2.1.6 compiled with icc 18.0.1
  252. OpenMPI 3.1.3 and 4.0.0 compiled with GCC 7.2.0
  253. PGI C, Fortran, C++ for 64-bit target on
  254. x86_64;
  255. Version 19.10-0
  256. Linux-3.10.0-1127.0.0.1chaos openmpi-4.0.0
  257. #1 SMP x86_64 GNU/Linux clang 6.0.0, 11.0.1
  258. (quartz) GCC 7.3.0, 8.1.0
  259. Intel 16.0.4, 18.0.2, 19.0.4
  260. macOS Apple M1 11.6 Apple clang version 12.0.5 (clang-1205.0.22.11)
  261. Darwin 20.6.0 arm64 gfortran GNU Fortran (Homebrew GCC 11.2.0) 11.1.0
  262. (macmini-m1) Intel icc/icpc/ifort version 2021.3.0 202106092021.3.0 20210609
  263. macOS Big Sur 11.3.1 Apple clang version 12.0.5 (clang-1205.0.22.9)
  264. Darwin 20.4.0 x86_64 gfortran GNU Fortran (Homebrew GCC 10.2.0_3) 10.2.0
  265. (bigsur-1) Intel icc/icpc/ifort version 2021.2.0 20210228
  266. macOS High Sierra 10.13.6 Apple LLVM version 10.0.0 (clang-1000.10.44.4)
  267. 64-bit gfortran GNU Fortran (GCC) 6.3.0
  268. (bear) Intel icc/icpc/ifort version 19.0.4.233 20190416
  269. macOS Sierra 10.12.6 Apple LLVM version 9.0.0 (clang-900.39.2)
  270. 64-bit gfortran GNU Fortran (GCC) 7.4.0
  271. (kite) Intel icc/icpc/ifort version 17.0.2
  272. Mac OS X El Capitan 10.11.6 Apple clang version 7.3.0 from Xcode 7.3
  273. 64-bit gfortran GNU Fortran (GCC) 5.2.0
  274. (osx1011test) Intel icc/icpc/ifort version 16.0.2
  275. Linux 2.6.32-573.22.1.el6 GNU C (gcc), Fortran (gfortran), C++ (g++)
  276. #1 SMP x86_64 GNU/Linux compilers:
  277. Centos6 Version 4.4.7 20120313
  278. (platypus) Version 4.9.3, 5.3.0, 6.2.0
  279. MPICH 3.1.4 compiled with GCC 4.9.3
  280. PGI C, Fortran, C++ for 64-bit target on
  281. x86_64;
  282. Version 19.10-0
  283. Windows 10 x64 Visual Studio 2015 w/ Intel C/C++/Fortran 18 (cmake)
  284. Visual Studio 2017 w/ Intel C/C++/Fortran 19 (cmake)
  285. Visual Studio 2019 w/ clang 12.0.0
  286. with MSVC-like command-line (C/C++ only - cmake)
  287. Visual Studio 2019 w/ Intel Fortran 19 (cmake)
  288. Visual Studio 2019 w/ MSMPI 10.1 (C only - cmake)
  289. Known Problems
  290. ==============
  291. Setting a variable-length dataset fill value will leak the memory allocated
  292. for the p field of the hvl_t struct. A fix is in progress for this.
  293. HDFFV-10840
  294. CMake files do not behave correctly with paths containing spaces.
  295. Do not use spaces in paths because the required escaping for handling spaces
  296. results in very complex and fragile build files.
  297. ADB - 2019/05/07
  298. At present, metadata cache images may not be generated by parallel
  299. applications. Parallel applications can read files with metadata cache
  300. images, but since this is a collective operation, a deadlock is possible
  301. if one or more processes do not participate.
  302. CPP ptable test fails on both VS2017 and VS2019 with Intel compiler, JIRA
  303. issue: HDFFV-10628. This test will pass with VS2015 with Intel compiler.
  304. The subsetting option in ph5diff currently will fail and should be avoided.
  305. The subsetting option works correctly in serial h5diff.
  306. Known problems in previous releases can be found in the HISTORY*.txt files
  307. in the HDF5 source. Please report any new problems found to
  308. help@hdfgroup.org.
  309. CMake vs. Autotools installations
  310. =================================
  311. While both build systems produce similar results, there are differences.
  312. Each system produces the same set of folders on linux (only CMake works
  313. on standard Windows); bin, include, lib and share. Autotools places the
  314. COPYING and RELEASE.txt file in the root folder, CMake places them in
  315. the share folder.
  316. The bin folder contains the tools and the build scripts. Additionally, CMake
  317. creates dynamic versions of the tools with the suffix "-shared". Autotools
  318. installs one set of tools depending on the "--enable-shared" configuration
  319. option.
  320. build scripts
  321. -------------
  322. Autotools: h5c++, h5cc, h5fc
  323. CMake: h5c++, h5cc, h5hlc++, h5hlcc
  324. The include folder holds the header files and the fortran mod files. CMake
  325. places the fortran mod files into separate shared and static subfolders,
  326. while Autotools places one set of mod files into the include folder. Because
  327. CMake produces a tools library, the header files for tools will appear in
  328. the include folder.
  329. The lib folder contains the library files, and CMake adds the pkgconfig
  330. subfolder with the hdf5*.pc files used by the bin/build scripts created by
  331. the CMake build. CMake separates the C interface code from the fortran code by
  332. creating C-stub libraries for each Fortran library. In addition, only CMake
  333. installs the tools library. The names of the szip libraries are different
  334. between the build systems.
  335. The share folder will have the most differences because CMake builds include
  336. a number of CMake specific files for support of CMake's find_package and support
  337. for the HDF5 Examples CMake project.
  338. The issues with the gif tool are:
  339. HDFFV-10592 CVE-2018-17433
  340. HDFFV-10593 CVE-2018-17436
  341. HDFFV-11048 CVE-2020-10809
  342. These CVE issues have not yet been addressed and can be avoided by not building
  343. the gif tool. Disable building the High-Level tools with these options:
  344. autotools: --disable-hltools
  345. cmake: HDF5_BUILD_HL_TOOLS=OFF