Glibc 2.18 Supports New Optimizations, New Archs
Version 2.18 of glibc is available and with it comes a host of new features, new architectures are supported, and various bug-fixes.
The official glibc 2.18 announcement has yet to surface, but the 2.18 release has been tagged in Git (and glibc 2.19 development now open), and packages of it can be downloaded. Fortunately, in looking at the Git tree we can already talk about the goodies of glibc 2.18 without the official release announcement.
Glibc 2.18 has improved C++11 support, performance related improvements / optimized string functions, a benchmark framework for glibc functions, new API functions, support for Intel TSX lock elision, and a whole lot of bug-fixes.
New CPU architectures supported by glibc 2.18 are Xilinx MicroBlaze and IBM POWER8. While not new architectures, there's also various 32-bit ARM and AARch64 improvements worth noting.
For the lack of an official glibc 2.18 release announcement at this time, embedded below are the 2.18 news items as obtained via Git.
The official glibc 2.18 announcement has yet to surface, but the 2.18 release has been tagged in Git (and glibc 2.19 development now open), and packages of it can be downloaded. Fortunately, in looking at the Git tree we can already talk about the goodies of glibc 2.18 without the official release announcement.
Glibc 2.18 has improved C++11 support, performance related improvements / optimized string functions, a benchmark framework for glibc functions, new API functions, support for Intel TSX lock elision, and a whole lot of bug-fixes.
New CPU architectures supported by glibc 2.18 are Xilinx MicroBlaze and IBM POWER8. While not new architectures, there's also various 32-bit ARM and AARch64 improvements worth noting.
For the lack of an official glibc 2.18 release announcement at this time, embedded below are the 2.18 news items as obtained via Git.
- The following bugs are resolved with this release: 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120, 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988, 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496, 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003, 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062, 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234, 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335, 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381, 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424, 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485, 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577, 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711, 15755, 15759.
- CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal has been fixed by disabling the use of pt_chown (Bugzilla #15755). Distributions can re-enable building and using pt_chown via the new configure option `--enable-pt_chown'. Enabling the use of pt_chown carries with it considerable security risks and should only be used if the distribution understands and accepts the risks.
- CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla #15078).
- CVE-2013-1914 Stack overflow in getaddrinfo with many results has been fixed (Bugzilla #15330).
- Add support for calling C++11 thread_local object destructors on thread and program exit. This needs compiler support for offloading C++11 destructor calls to glibc.
- Improved worst case performance of libm functions with double inputs and output.
- Support for priority inherited mutexes in pthread condition variables on non-x86 architectures.
- Port to Xilinx MicroBlaze contributed by David Holsgrove.
- Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
- Optimized string functions for ARM. Implemented by Will Newton and Richard Henderson.
- Optimized libm functions for SPARC. Implemented by David S. Miller.
- Support for building more of ARM glibc as Thumb-2 code. Implemented by Richard Henderson.
- Support for building most of MIPS glibc as MIPS16 code. Implemented by Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
- Added a benchmark framework to track performance of functions in glibc.
- New