1 GNU C Library NEWS -- history of user-visible changes.
2 Copyright (C) 1992-2018 Free Software Foundation, Inc.
3 See the end for copying conditions.
5 Please send GNU C library bug reports via <https://sourceware.org/bugzilla/>
6 using `glibc' in the "product" field.
10 The following bugs are resolved with this release:
12 [19444] build failures with -O1 due to -Wmaybe-uninitialized
13 [20209] localedata: Spelling mistake for Sunday in Greenlandic kl_GL
14 [22927] libanl: properly cleanup if first helper thread creation failed
15 [23400] stdlib/test-bz22786.c creates temporary files in glibc source tree
16 [23497] readdir64@GLIBC_2.1 cannot parse the kernel directory stream
17 [23509] CET enabled glibc is incompatible with the older linker
18 [23521] nss_files aliases database file stream leak
19 [23538] pthread_cond_broadcast: Fix waiters-after-spinning case
20 [23562] signal: Use correct type for si_band in siginfo_t
21 [23578] regex: Fix memory overread in re_compile_pattern
22 [23579] libc: Errors misreported in preadv2
23 [23606] Missing ENDBR32 in sysdeps/i386/start.S
24 [23614] powerpc: missing CFI register information in __mpn_* functions
25 [23679] gethostid: Missing NULL check for gethostbyname_r result
26 [23709] Fix CPU string flags for Haswell-type CPUs
27 [23717] Fix stack overflow in stdlib/tst-setcontext9
28 [23821] si_band in siginfo_t has wrong type long int on sparc64
29 [23822] ia64 static libm.a is missing exp2f, log2f and powf symbols
30 [23927] Linux if_nametoindex() does not close descriptor (CVE-2018-19591)
31 [23972] __old_getdents64 uses wrong d_off value on overflow
32 [24018] gettext may return NULL
33 [24022] riscv may lack <asm/syscalls.h>
34 [24024] strerror() might set errno to ENOMEM due to -fno-math-error
35 [24027] malloc: Integer overflow in realloc
36 [24034] tst-cancel21-static fails with SIGBUS on pre-ARMv7 when using GCC 8
37 [24097] Can't use 64-bit register for size_t in assembly codes for x32 (CVE-2019-6488)
38 [24155] x32 memcmp can treat positive length as 0 (if sign bit in RDX is set) (CVE-2019-7309)
40 Security related changes:
42 CVE-2018-19591: A file descriptor leak in if_nametoindex can lead to a
43 denial of service due to resource exhaustion when processing getaddrinfo
44 calls with crafted host names. Reported by Guido Vranken.
46 CVE-2019-6488: On x32, the size_t parameter may be passed in the lower
47 32 bits of a 64-bit register with with non-zero upper 32 bit. When it
48 happened, accessing the 32-bit size_t value as the full 64-bit register
49 in the assembly string/memory functions would cause a buffer overflow.
52 CVE-2019-7309: x86-64 memcmp used signed Jcc instructions to check
53 size. For x86-64, memcmp on an object size larger than SSIZE_MAX
54 has undefined behavior. On x32, the size_t argument may be passed
55 in the lower 32 bits of the 64-bit RDX register with non-zero upper
56 32 bits. When it happened with the sign bit of RDX register set,
57 memcmp gave the wrong result since it treated the size argument as
58 zero. Reported by H.J. Lu.
65 * The localization data for ISO 14651 is updated to match the 2016
66 Edition 4 release of the standard, this matches data provided by
67 Unicode 9.0.0. This update introduces significant improvements to the
68 collation of Unicode characters. This release deviates slightly from
69 the standard in that the collation element ordering for lowercase and
70 uppercase LATIN script characters is adjusted to ensure that regular
71 expressions with ranges like [a-z] and [A-Z] don't interleave e.g. A
72 is not matched by [a-z]. With the update many locales have been
73 updated to take advantage of the new collation information. The new
74 collation information has increased the size of the compiled locale
75 archive or binary locales.
77 * The GNU C Library can now be compiled with support for Intel CET, AKA
78 Intel Control-flow Enforcement Technology. When the library is built
79 with --enable-cet, the resulting glibc is protected with indirect
80 branch tracking (IBT) and shadow stack (SHSTK). CET-enabled glibc is
81 compatible with all existing executables and shared libraries. This
82 feature is currently supported on i386, x86_64 and x32 with GCC 8 and
83 binutils 2.29 or later. Note that CET-enabled glibc requires CPUs
84 capable of multi-byte NOPs, like x86-64 processors as well as Intel
85 Pentium Pro or newer. NOTE: --enable-cet has been tested for i686,
86 x86_64 and x32 on non-CET processors. --enable-cet has been tested
87 for x86_64 and x32 on CET SDVs, but Intel CET support hasn't been
90 * The GNU C Library now has correct support for ABSOLUTE symbols
91 (SHN_ABS-relative symbols). Previously such ABSOLUTE symbols were
92 relocated incorrectly or in some cases discarded. The GNU linker can
93 make use of the newer semantics, but it must communicate it to the
94 dynamic loader by setting the ELF file's identification (EI_ABIVERSION
95 field) to indicate such support is required.
97 * Unicode 11.0.0 Support: Character encoding, character type info, and
98 transliteration tables are all updated to Unicode 11.0.0, using
99 generator scripts contributed by Mike FABIAN (Red Hat).
101 * <math.h> functions that round their results to a narrower type are added
102 from TS 18661-1:2014 and TS 18661-3:2015:
104 - fadd, faddl, daddl and corresponding fMaddfN, fMaddfNx, fMxaddfN and
107 - fsub, fsubl, dsubl and corresponding fMsubfN, fMsubfNx, fMxsubfN and
110 - fmul, fmull, dmull and corresponding fMmulfN, fMmulfNx, fMxmulfN and
113 - fdiv, fdivl, ddivl and corresponding fMdivfN, fMdivfNx, fMxdivfN and
116 * Two grammatical forms of month names are now supported for the following
117 languages: Armenian, Asturian, Catalan, Czech, Kashubian, Occitan, Ossetian,
118 Scottish Gaelic, Upper Sorbian, and Walloon. The following languages now
119 support two grammatical forms in abbreviated month names: Catalan, Greek,
122 * Newly added locales: Lower Sorbian (dsb_DE) and Yakut (sah_RU) also
123 include the support for two grammatical forms of month names.
125 * Building and running on GNU/Hurd systems now works without out-of-tree
128 * The renameat2 function has been added, a variant of the renameat function
129 which has a flags argument. If the flags are zero, the renameat2 function
130 acts like renameat. If the flag is not zero and there is no kernel
131 support for renameat2, the function will fail with an errno value of
132 EINVAL. This is different from the existing gnulib function renameatu,
133 which performs a plain rename operation in case of a RENAME_NOREPLACE
134 flags and a non-existing destination (and therefore has a race condition
135 that can clobber the destination inadvertently).
137 * The statx function has been added, a variant of the fstatat64
138 function with an additional flags argument. If there is no direct
139 kernel support for statx, glibc provides basic stat support based on
140 the fstatat64 function.
142 * IDN domain names in getaddrinfo and getnameinfo now use the system libidn2
143 library if installed. libidn2 version 2.0.5 or later is recommended. If
144 libidn2 is not available, internationalized domain names are not encoded
145 or decoded even if the AI_IDN or NI_IDN flags are passed to getaddrinfo or
146 getnameinfo. (getaddrinfo calls with non-ASCII names and AI_IDN will fail
147 with an encoding error.) Flags which used to change the IDN encoding and
148 decoding behavior (AI_IDN_ALLOW_UNASSIGNED, AI_IDN_USE_STD3_ASCII_RULES,
149 NI_IDN_ALLOW_UNASSIGNED, NI_IDN_USE_STD3_ASCII_RULES) have been
150 deprecated. They no longer have any effect.
152 * Parsing of dynamic string tokens in DT_RPATH, DT_RUNPATH, DT_NEEDED,
153 DT_AUXILIARY, and DT_FILTER has been expanded to support the full
154 range of ELF gABI expressions including such constructs as
155 '$ORIGIN$ORIGIN' (if valid). For SUID/GUID applications the rules
156 have been further restricted, and where in the past a dynamic string
157 token sequence may have been interpreted as a literal string it will
158 now cause a load failure. These load failures were always considered
159 unspecified behaviour from the perspective of the dynamic loader, and
160 for safety are now load errors e.g. /foo/${ORIGIN}.so in DT_NEEDED
161 results in a load failure now.
163 * Support for ISO C threads (ISO/IEC 9899:2011) has been added. The
164 implementation includes all the standard functions provided by
167 - thrd_current, thrd_equal, thrd_sleep, thrd_yield, thrd_create,
168 thrd_detach, thrd_exit, and thrd_join for thread management.
170 - mtx_init, mtx_lock, mtx_timedlock, mtx_trylock, mtx_unlock, and
171 mtx_destroy for mutual exclusion.
173 - call_once for function call synchronization.
175 - cnd_broadcast, cnd_destroy, cnd_init, cnd_signal, cnd_timedwait, and
176 cnd_wait for conditional variables.
178 - tss_create, tss_delete, tss_get, and tss_set for thread-local storage.
180 Application developers must link against libpthread to use ISO C threads.
182 Deprecated and removed features, and other changes affecting compatibility:
184 * The nonstandard header files <libio.h> and <_G_config.h> are no longer
185 installed. Software that was using either header should be updated to
186 use standard <stdio.h> interfaces instead.
188 * The stdio functions 'getc' and 'putc' are no longer defined as macros.
189 This was never required by the C standard, and the macros just expanded
190 to call alternative names for the same functions. If you hoped getc and
191 putc would provide performance improvements over fgetc and fputc, instead
192 investigate using (f)getc_unlocked and (f)putc_unlocked, and, if
193 necessary, flockfile and funlockfile.
195 * All stdio functions now treat end-of-file as a sticky condition. If you
196 read from a file until EOF, and then the file is enlarged by another
197 process, you must call clearerr or another function with the same effect
198 (e.g. fseek, rewind) before you can read the additional data. This
199 corrects a longstanding C99 conformance bug. It is most likely to affect
200 programs that use stdio to read interactive input from a terminal.
203 * The macros 'major', 'minor', and 'makedev' are now only available from
204 the header <sys/sysmacros.h>; not from <sys/types.h> or various other
205 headers that happen to include <sys/types.h>. These macros are rarely
206 used, not part of POSIX nor XSI, and their names frequently collide with
207 user code; see https://sourceware.org/bugzilla/show_bug.cgi?id=19239 for
210 <sys/sysmacros.h> is a GNU extension. Portable programs that require
211 these macros should first include <sys/types.h>, and then include
212 <sys/sysmacros.h> if __GNU_LIBRARY__ is defined.
214 * The tilegx*-*-linux-gnu configurations are no longer supported.
216 * The obsolete function ustat is no longer available to newly linked
217 binaries; the headers <ustat.h> and <sys/ustat.h> have been removed. This
218 function has been deprecated in favor of fstatfs and statfs.
220 * The obsolete function nfsservctl is no longer available to newly linked
221 binaries. This function was specific to systems using the Linux kernel
222 and could not usefully be used with the GNU C Library on systems with
223 version 3.1 or later of the Linux kernel.
225 * The obsolete function name llseek is no longer available to newly linked
226 binaries. This function was specific to systems using the Linux kernel
227 and was not declared in a header. Programs should use the lseek64 name
228 for this function instead.
230 * The AI_IDN_ALLOW_UNASSIGNED and NI_IDN_ALLOW_UNASSIGNED flags for the
231 getaddrinfo and getnameinfo functions have been deprecated. The behavior
232 previously selected by them is now always enabled.
234 * The AI_IDN_USE_STD3_ASCII_RULES and NI_IDN_USE_STD3_ASCII_RULES flags for
235 the getaddrinfo and getnameinfo functions have been deprecated. The STD3
236 restriction (rejecting '_' in host names, among other things) has been
237 removed, for increased compatibility with non-IDN name resolution.
239 * The fcntl function now have a Long File Support variant named fcntl64. It
240 is added to fix some Linux Open File Description (OFD) locks usage on non
241 LFS mode. As for others *64 functions, fcntl64 semantics are analogous with
242 fcntl and LFS support is handled transparently. Also for Linux, the OFD
243 locks act as a cancellation entrypoint.
245 * The obsolete functions encrypt, encrypt_r, setkey, setkey_r, cbc_crypt,
246 ecb_crypt, and des_setparity are no longer available to newly linked
247 binaries, and the headers <rpc/des_crypt.h> and <rpc/rpc_des.h> are no
248 longer installed. These functions encrypted and decrypted data with the
249 DES block cipher, which is no longer considered secure. Software that
250 still uses these functions should switch to a modern cryptography library,
253 * Reflecting the removal of the encrypt and setkey functions above, the
254 macro _XOPEN_CRYPT is no longer defined. As a consequence, the crypt
255 function is no longer declared unless _DEFAULT_SOURCE or _GNU_SOURCE is
258 * The obsolete function fcrypt is no longer available to newly linked
259 binaries. It was just another name for the standard function crypt,
260 and it has not appeared in any header file in many years.
262 * We have tentative plans to hand off maintenance of the passphrase-hashing
263 library, libcrypt, to a separate development project that will, we hope,
264 keep up better with new passphrase-hashing algorithms. We will continue
265 to declare 'crypt' in <unistd.h>, and programs that use 'crypt' or
266 'crypt_r' should not need to change at all; however, distributions will
267 need to install <crypt.h> and libcrypt from a separate project.
269 In this release, if the configure option --disable-crypt is used, glibc
270 will not install <crypt.h> or libcrypt, making room for the separate
271 project's versions of these files. The plan is to make this the default
272 behavior in a future release.
274 Changes to build and runtime requirements:
276 GNU make 4.0 or later is now required to build glibc.
278 Security related changes:
280 CVE-2016-6261, CVE-2016-6263, CVE-2017-14062: Various vulnerabilities have
281 been fixed by removing the glibc-internal IDNA implementation and using
282 the system-provided libidn2 library instead. Originally reported by Hanno
283 Böck and Christian Weisgerber.
285 CVE-2017-18269: An SSE2-based memmove implementation for the i386
286 architecture could corrupt memory. Reported by Max Horn.
288 CVE-2018-11236: Very long pathname arguments to realpath function could
289 result in an integer overflow and buffer overflow. Reported by Alexey
292 CVE-2018-11237: The mempcpy implementation for the Intel Xeon Phi
293 architecture could write beyond the target buffer, resulting in a buffer
294 overflow. Reported by Andreas Schwab.
296 The following bugs are resolved with this release:
298 [1190] stdio: fgetc()/fread() behaviour is not POSIX compliant
299 [6889] manual: 'PWD' mentioned but not specified
300 [13575] libc: SSIZE_MAX defined as LONG_MAX is inconsistent with ssize_t,
301 when __WORDSIZE != 64
302 [13762] regex: re_search etc. should return -2 on memory exhaustion
303 [13888] build: /tmp usage during testing
304 [13932] math: dbl-64 pow unexpectedly slow for some inputs
305 [14092] nptl: Support C11 threads
306 [14095] localedata: Review / update collation data from Unicode / ISO
308 [14508] libc: -Wformat warnings
309 [14553] libc: Namespace pollution loff_t in sys/types.h
310 [14890] libc: Make NT_PRFPREG canonical.
311 [15105] libc: Extra PLT references with -Os
312 [15512] libc: __bswap_constant_16 not compiled when -Werror -Wsign-
314 [16335] manual: Feature test macro documentation incomplete and out of
316 [16552] libc: Unify umount implementations in terms of umount2
317 [17082] libc: htons et al.: statement-expressions prevent use on global
318 scope with -O1 and higher
319 [17343] libc: Signed integer overflow in /stdlib/random_r.c
320 [17438] localedata: pt_BR: wrong d_fmt delimiter
321 [17662] libc: please implement binding for the new renameat2 syscall
322 [17721] libc: __restrict defined as /* Ignore */ even in c11
323 [17979] libc: inconsistency between uchar.h and stdint.h
324 [18018] dynamic-link: Additional $ORIGIN handling issues (CVE-2011-0536)
325 [18023] libc: extend_alloca is broken (questionable pointer comparison,
326 horrible machine code)
327 [18124] libc: hppa: setcontext erroneously returns -1 as exit code for
329 [18471] libc: llseek should be a compat symbol
330 [18473] soft-fp: [powerpc-nofpu] __sqrtsf2, __sqrtdf2 should be compat
332 [18991] nss: nss_files skips large entry in database
333 [19239] libc: Including stdlib.h ends up with macros major and minor being
335 [19463] libc: linknamespace failures when compiled with -Os
336 [19485] localedata: csb_PL: Update month translations + add yesstr/nostr
337 [19527] locale: Normalized charset name not recognized by setlocale
338 [19667] string: Missing Sanity Check for malloc calls in file 'testcopy.c'
339 [19668] libc: Missing Sanity Check for malloc() in file 'tst-setcontext-
341 [19728] network: out of bounds stack read in libidn function
342 idna_to_ascii_4i (CVE-2016-6261)
343 [19729] network: out of bounds heap read on invalid utf-8 inputs in
344 stringprep_utf8_nfkc_normalize (CVE-2016-6263)
345 [19818] dynamic-link: Absolute (SHN_ABS) symbols incorrectly relocated by
347 [20079] libc: Add SHT_X86_64_UNWIND to elf.h
348 [20251] libc: 32bit programs pass garbage in struct flock for OFD locks
349 [20419] dynamic-link: files with large allocated notes crash in
351 [20530] libc: bswap_16 should use __builtin_bswap16() when available
352 [20890] dynamic-link: ldconfig: fsync the files before atomic rename
353 [20980] manual: CFLAGS environment variable replaces vital options
354 [21163] regex: Assertion failure in pop_fail_stack when executing a
355 malformed regexp (CVE-2015-8985)
356 [21234] manual: use of CFLAGS makes glibc detect no optimization
357 [21269] dynamic-link: i386 sigaction sa_restorer handling is wrong
358 [21313] build: Compile Error GCC 5.4.0 MIPS with -0S
359 [21314] build: Compile Error GCC 5.2.0 MIPS with -0s
360 [21508] locale: intl/tst-gettext failure with latest msgfmt
361 [21547] localedata: Tibetan script collation broken (Dzongkha and Tibetan)
362 [21812] network: getifaddrs() returns entries with ifa_name == NULL
363 [21895] libc: ppc64 setjmp/longjmp not fully interoperable with static
365 [21942] dynamic-link: _dl_dst_substitute incorrectly handles $ORIGIN: with
367 [22241] localedata: New locale: Yakut (Sakha) locale for Russia (sah_RU)
368 [22247] network: Integer overflow in the decode_digit function in
369 puny_decode.c in libidn (CVE-2017-14062)
370 [22342] nscd: NSCD not properly caching netgroup
371 [22391] nptl: Signal function clear NPTL internal symbols inconsistently
372 [22550] localedata: es_ES locale (and other es_* locales): collation
373 should treat ñ as a primary different character, sync the collation
374 for Spanish with CLDR
375 [22638] dynamic-link: sparc: static binaries are broken if glibc is built
376 by gcc configured with --enable-default-pie
377 [22639] time: year 2039 bug for localtime etc. on 64-bit platforms
378 [22644] string: memmove-sse2-unaligned on 32bit x86 produces garbage when
379 crossing 2GB threshold (CVE-2017-18269)
380 [22646] localedata: redundant data (LC_TIME) for es_CL, es_CU, es_EC and
382 [22735] time: Misleading typo in time.h source comment regarding
384 [22753] libc: preadv2/pwritev2 fallback code should handle offset=-1
385 [22761] libc: No trailing `%n' conversion specifier in FMT passed from
386 `__assert_perror_fail ()' to `__assert_fail_base ()'
387 [22766] libc: all glibc internal dlopen should use RTLD_NOW for robust
389 [22786] libc: Stack buffer overflow in realpath() if input size is close
390 to SSIZE_MAX (CVE-2018-11236)
391 [22787] dynamic-link: _dl_check_caller returns false when libc is linked
392 through an absolute DT_NEEDED path
393 [22792] build: tcb-offsets.h dependency dropped
394 [22797] libc: pkey_get() uses non-reserved name of argument
395 [22807] libc: PTRACE_* constants missing for powerpc
396 [22818] glob: posix/tst-glob_lstat_compat failure on alpha
397 [22827] dynamic-link: RISC-V ELF64 parser mis-reads flag in ldconfig
398 [22830] malloc: malloc_stats doesn't restore cancellation state on stderr
399 [22848] localedata: ca_ES: update date definitions from CLDR
400 [22862] build: _DEFAULT_SOURCE is defined even when _ISOC11_SOURCE is
401 [22884] math: RISCV fmax/fmin handle signalling NANs incorrectly
402 [22896] localedata: Update locale data for an_ES
403 [22902] math: float128 test failures with GCC 8
404 [22918] libc: multiple common of `__nss_shadow_database'
405 [22919] libc: sparc32: backtrace yields infinite backtrace with
407 [22926] libc: FTBFS on powerpcspe
408 [22932] localedata: lt_LT: Update of abbreviated month names from CLDR
410 [22937] localedata: Greek (el_GR, el_CY) locales actually need ab_alt_mon
411 [22947] libc: FAIL: misc/tst-preadvwritev2
412 [22963] localedata: cs_CZ: Add alternative month names
413 [22987] math: [powerpc/sparc] fdim inlines errno, exceptions handling
414 [22996] localedata: change LC_PAPER to en_US in es_BO locale
415 [22998] dynamic-link: execstack tests are disabled when SELinux is
417 [23005] network: Crash in __res_context_send after memory allocation
419 [23007] math: strtod cannot handle -nan
420 [23024] nss: getlogin_r is performing NSS lookups when loginid isn't set
421 [23036] regex: regex equivalence class regression
422 [23037] libc: initialize msg_flags to zero for sendmmsg() calls
423 [23069] libc: sigaction broken on riscv64-linux-gnu
424 [23094] localedata: hr_HR: wrong thousands_sep and mon_thousands_sep
425 [23102] dynamic-link: Incorrect parsing of multiple consecutive $variable
426 patterns in runpath entries (e.g. $ORIGIN$ORIGIN)
427 [23137] nptl: s390: pthread_join sometimes block indefinitely (on 31bit
428 and libc build with -Os)
429 [23140] localedata: More languages need two forms of month names
430 [23145] libc: _init/_fini aren't marked as hidden
431 [23152] localedata: gd_GB: Fix typo in "May" (abbreviated)
432 [23171] math: C++ iseqsig for long double converts arguments to double
433 [23178] nscd: sudo will fail when it is run in concurrent with commands
434 that changes /etc/passwd
435 [23196] string: __mempcpy_avx512_no_vzeroupper mishandles large copies
437 [23206] dynamic-link: static-pie + dlopen breaks debugger interaction
438 [23208] localedata: New locale - Lower Sorbian (dsb)
439 [23233] regex: Memory leak in build_charclass_op function in file
441 [23236] stdio: Harden function pointers in _IO_str_fields
442 [23250] nptl: Offset of __private_ss differs from GCC
443 [23253] math: tgamma test suite failures on i686 with -march=x86-64
444 -mtune=generic -mfpmath=sse
445 [23259] dynamic-link: Unsubstituted ${ORIGIN} remains in DT_NEEDED for
447 [23264] libc: posix_spawnp wrongly executes ENOEXEC in non compat mode
448 [23266] nis: stringop-truncation warning with new gcc8.1 in nisplus-
450 [23272] math: fma(INFINITY,INFIITY,0.0) should be INFINITY
451 [23277] math: nan function should not have const attribute
452 [23279] math: scanf and strtod wrong for some hex floating-point
453 [23280] math: wscanf rounds wrong; wcstod is ok for negative numbers and
455 [23290] localedata: IBM273 is not equivalent to ISO-8859-1
456 [23303] build: undefined reference to symbol
457 '__parse_hwcap_and_convert_at_platform@@GLIBC_2.23'
458 [23307] dynamic-link: Absolute symbols whose value is zero ignored in
460 [23313] stdio: libio vtables validation and standard file object
462 [23329] libc: The __libc_freeres infrastructure is not properly run across
464 [23349] libc: Various glibc headers no longer compatible with
466 [23351] malloc: Remove unused code related to heap dumps and malloc
468 [23363] stdio: stdio-common/tst-printf.c has non-free license
469 [23396] regex: Regex equivalence regression in single-byte locales
470 [23422] localedata: oc_FR: More updates of locale data
471 [23442] build: New warning with GCC 8
472 [23448] libc: Out of bounds access in IBM-1390 converter
473 [23456] libc: Wrong index_cpu_LZCNT
474 [23458] build: tst-get-cpu-features-static isn't added to tests
475 [23459] libc: COMMON_CPUID_INDEX_80000001 isn't populated for Intel
477 [23467] dynamic-link: x86/CET: A property note parser bug
484 * The GNU C Library can now be compiled with support for building static
485 PIE executables (See --enable-static-pie in INSTALL). These static PIE
486 executables are like static executables but can be loaded at any address
487 and provide additional security hardening benefits at the cost of some
488 memory and performance. When the library is built with --enable-static-pie
489 the resulting libc.a is usable with GCC 8 and above to create static PIE
490 executables using the GCC option '-static-pie'. This feature is currently
491 supported on i386, x86_64 and x32 with binutils 2.29 or later, and on
492 aarch64 with binutils 2.30 or later.
494 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf,
495 sinf, sincosf and tan with FMA, contributed by Arjan van de Ven and
498 * Optimized x86-64 trunc and truncf for processors with SSE4.1.
500 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf.
502 * In order to support faster and safer process termination the malloc API
503 family of functions will no longer print a failure address and stack
504 backtrace after detecting heap corruption. The goal is to minimize the
505 amount of work done after corruption is detected and to avoid potential
506 security issues in continued process execution. Reducing shutdown time
507 leads to lower overall process restart latency, so there is benefit both
508 from a security and performance perspective.
510 * The abort function terminates the process immediately, without flushing
511 stdio streams. Previous glibc versions used to flush streams, resulting
512 in deadlocks and further data corruption. This change also affects
513 process aborts as the result of assertion failures.
515 * On platforms where long double has the IEEE binary128 format (aarch64,
516 alpha, mips64, riscv, s390 and sparc), the math library now implements
517 _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015.
518 These are the same interfaces added in version 2.26 for some platforms where
519 this format is supported but is not the format of long double.
521 * On platforms with support for _Float64x (aarch64, alpha, i386, ia64,
522 mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now
523 implements interfaces for that type, as defined by ISO/IEC TS
524 18661-3:2015. These are corresponding interfaces to those supported for
527 * The math library now implements interfaces for the _Float32, _Float64 and
528 _Float32x types, as defined by ISO/IEC TS 18661-3:2015. These are
529 corresponding interfaces to those supported for _Float128.
531 * glibc now implements the memfd_create and mlock2 functions on Linux.
533 * Support for memory protection keys was added. The <sys/mman.h> header now
534 declares the functions pkey_alloc, pkey_free, pkey_mprotect, pkey_set,
537 * The copy_file_range function was added.
539 * Optimized memcpy, mempcpy, memmove, and memset for sparc M7.
541 * The ldconfig utility now processes `include' directives using the C/POSIX
542 collation ordering. Previous glibc versions used locale-specific
543 ordering, the change might break systems that relied on that.
545 * Support for two grammatical forms of month names has been added.
546 In a call to strftime, the "%B" and "%b" format specifiers will now
547 produce the grammatical form required when the month is used as part
548 of a complete date. New "%OB" and "%Ob" specifiers produce the form
549 required when the month is named by itself. For instance, in Greek
550 and in many Slavic and Baltic languages, "%B" will produce the month
551 in genitive case, and "%OB" will produce the month in nominative case.
553 In a call to strptime, "%B", "%b", "%h", "%OB", "%Ob", and "%Oh"
554 are all valid and will all accept any known form of month
555 name---standalone or complete, abbreviated or full. In a call to
556 nl_langinfo, the query constants MON_1..12 and ABMON_1..12 return
557 the strings used by "%B" and "%b", respectively. New query
558 constants ALTMON_1..12 and _NL_ABALTMON_1..12 return the strings
559 used by "%OB" and "%Ob", respectively.
561 In a locale definition file, use "alt_mon" and "ab_alt_mon" to
562 define the strings for %OB and %Ob, respectively; these have the
563 same syntax as "mon" and "abmon". These arrays are optional; if they
564 are not provided then they have the same content as "mon" and "abmon",
567 These features are provided for locales which define "alt_mon" and/or
568 "ab_alt_mon" in their locale source data. This release includes such
569 alternative month name data for the following languages: Belarusian,
570 Croatian, Greek, Lithuanian, Polish, Russian, and Ukrainian.
572 This feature is currently a GNU extension, but it is expected to
573 be added to the next revision of POSIX, and it is also already
574 available on some BSD-derived operating systems.
576 This feature will cause existing statically compiled applications
577 to fail to load locales and fall back to the builtin C/POSIX locales.
578 See notes below for other changes affecting compatibility.
580 * Support for the RISC-V ISA running on Linux has been added. This port
581 requires at least binutils-2.30, gcc-7.3.0, and linux-4.15; and is supported
582 for the following ISA and ABI pairs:
588 Deprecated and removed features, and other changes affecting compatibility:
590 * Statically compiled applications attempting to load locales compiled for the
591 GNU C Library version 2.27 will fail and fall back to the builtin C/POSIX
592 locale. The reason for this is that the addition of the new "%OB" and "%Ob",
593 support for two grammatical forms of the month names, also extends the locale
594 data binary format. Static applications needing locale support must be
595 recompiled to match the runtime and data they are deployed with. In some
596 distributions there is an upgrade window where dynamically linked applications
597 may use a new library but the old locale data and also fall back to the
598 builtin C/POSIX locales; restarting the application process is sufficient to
601 * Support for statically linked applications which call dlopen is deprecated
602 and will be removed in a future version of glibc. Applications which call
603 dlopen need to be linked dynamically instead.
605 * Support for old programs which use internal stdio data structures and
606 functions is deprecated. This includes programs which use the C++ streams
607 provided by libstdc++ in GCC 2.95. Programs which use the internal
608 symbols _IO_adjust_wcolumn, _IO_default_doallocate, _IO_default_finish,
609 _IO_default_pbackfail, _IO_default_uflow, _IO_default_xsgetn,
610 _IO_default_xsputn, _IO_doallocbuf, _IO_do_write, _IO_file_attach,
611 _IO_file_close, _IO_file_close_it, _IO_file_doallocate, _IO_file_fopen,
612 _IO_file_init, _IO_file_jumps, _IO_fileno, _IO_file_open,
613 _IO_file_overflow, _IO_file_read, _IO_file_seek, _IO_file_seekoff,
614 _IO_file_setbuf, _IO_file_stat, _IO_file_sync, _IO_file_underflow,
615 _IO_file_write, _IO_file_xsputn, _IO_flockfile, _IO_flush_all,
616 _IO_flush_all_linebuffered, _IO_free_backup_area, _IO_free_wbackup_area,
617 _IO_init, _IO_init_marker, _IO_init_wmarker, _IO_iter_begin, _IO_iter_end,
618 _IO_iter_file, _IO_iter_next, _IO_least_wmarker, _IO_link_in,
619 _IO_list_all, _IO_list_lock, _IO_list_resetlock, _IO_list_unlock,
620 _IO_marker_delta, _IO_marker_difference, _IO_remove_marker, _IO_seekmark,
621 _IO_seekwmark, _IO_str_init_readonly, _IO_str_init_static,
622 _IO_str_overflow, _IO_str_pbackfail, _IO_str_seekoff, _IO_str_underflow,
623 _IO_switch_to_main_wget_area, _IO_switch_to_wget_mode,
624 _IO_unsave_wmarkers, _IO_wdefault_doallocate, _IO_wdefault_finish,
625 _IO_wdefault_pbackfail, _IO_wdefault_setbuf, _IO_wdefault_uflow,
626 _IO_wdefault_xsgetn, _IO_wdefault_xsputn, _IO_wdoallocbuf, _IO_wdo_write,
627 _IO_wfile_jumps, _IO_wfile_overflow, _IO_wfile_sync, _IO_wfile_underflow,
628 _IO_wfile_xsputn, _IO_wmarker_delta, or _IO_wsetb may stop working with a
629 future version of glibc. Unlike other symbol removals, these old
630 applications will not be supported using compatibility symbols.
632 * On GNU/Linux, the obsolete Linux constant PTRACE_SEIZE_DEVEL is no longer
633 defined by <sys/ptrace.h>.
635 * libm no longer supports SVID error handling (calling a user-provided
636 matherr function on error) or the _LIB_VERSION variable to control error
637 handling. (SVID error handling and the _LIB_VERSION variable still work
638 for binaries linked against older versions of the GNU C Library.) The
639 libieee.a library is no longer provided. math.h no longer defines struct
640 exception, or the macros X_TLOSS, DOMAIN, SING, OVERFLOW, UNDERFLOW,
641 TLOSS, PLOSS and HUGE.
643 * The libm functions pow10, pow10f and pow10l are no longer supported for
644 new programs. Programs should use the standard names exp10, exp10f and
645 exp10l for these functions instead.
647 * The mcontext_t type is no longer the same as struct sigcontext. On
648 platforms where it was previously the same, this changes the C++ name
649 mangling for interfaces involving this type.
651 * The add-ons mechanism for building additional packages at the same time as
652 glibc has been removed. The --enable-add-ons configure option is now
655 * The --without-fp configure option is now ignored. Whether hardware
656 floating-point instructions are used is now configured based on whether
657 the compiler used at configure time (without any options implied by a
658 --with-cpu= configure option) uses such instructions.
660 * The res_hnok, res_dnok, res_mailok and res_ownok functions now check that
661 the specified string can be parsed as a domain name.
663 * In the malloc_info output, the <heap> element may contain another <aspace>
664 element, "subheaps", which contains the number of sub-heaps.
666 * The libresolv function p_secstodate is no longer supported for new
669 * The tilepro-*-linux-gnu configuration is no longer supported.
671 * The nonstandard header files <libio.h> and <_G_config.h> are deprecated
672 and will be removed in a future release. Software that is still using
673 either header should be updated to use standard <stdio.h> interfaces
676 libio.h was originally the header for a set of supported GNU extensions,
677 but they have not been maintained as such in many years, they are now
678 standing in the way of improvements to stdio, and we don't think there are
679 any remaining external users. _G_config.h was never intended for public
680 use, but predates the bits convention.
682 Changes to build and runtime requirements:
684 * bison version 2.7 or later is required to generate code in the 'intl'
687 Security related changes:
689 CVE-2009-5064: The ldd script would sometimes run the program under
690 examination directly, without preventing code execution through the
691 dynamic linker. (The glibc project disputes that this is a security
692 vulnerability; only trusted binaries must be examined using the ldd
695 CVE-2017-15670: The glob function, when invoked with GLOB_TILDE,
696 suffered from a one-byte overflow during ~ operator processing (either
697 on the stack or the heap, depending on the length of the user name).
698 Reported by Tim Rühsen.
700 CVE-2017-15671: The glob function, when invoked with GLOB_TILDE,
701 would sometimes fail to free memory allocated during ~ operator
702 processing, leading to a memory leak and, potentially, to a denial
705 CVE-2017-15804: The glob function, when invoked with GLOB_TILDE and
706 without GLOB_NOESCAPE, could write past the end of a buffer while
707 unescaping user names. Reported by Tim Rühsen.
709 CVE-2017-17426: The malloc function, when called with an object size near
710 the value SIZE_MAX, would return a pointer to a buffer which is too small,
711 instead of NULL. This was a regression introduced with the new malloc
712 thread cache in glibc 2.26. Reported by Iain Buclaw.
714 CVE-2017-1000408: Incorrect array size computation in _dl_init_paths leads
715 to the allocation of too much memory. (This is not a security bug per se,
716 it is mentioned here only because of the CVE assignment.) Reported by
719 CVE-2017-1000409: Buffer overflow in _dl_init_paths due to miscomputation
720 of the number of search path components. (This is not a security
721 vulnerability per se because no trust boundary is crossed if the fix for
722 CVE-2017-1000366 has been applied, but it is mentioned here only because
723 of the CVE assignment.) Reported by Qualys.
725 CVE-2017-16997: Incorrect handling of RPATH or RUNPATH containing $ORIGIN
726 for AT_SECURE or SUID binaries could be used to load libraries from the
729 CVE-2018-1000001: Buffer underflow in realpath function when getcwd function
730 succeeds without returning an absolute path due to unexpected behaviour
731 of the Linux kernel getcwd syscall. Reported by halfdog.
733 CVE-2018-6485: The posix_memalign and memalign functions, when called with
734 an object size near the value of SIZE_MAX, would return a pointer to a
735 buffer which is too small, instead of NULL. Reported by Jakub Wilk.
737 CVE-2018-6551: The malloc function, when called with an object size near
738 the value of SIZE_MAX, would return a pointer to a buffer which is too
739 small, instead of NULL.
741 The following bugs are resolved with this release:
743 [866] glob: glob should match dangling symlinks
744 [1062] glob: sysdeps/generic/glob.c merge from gnulib (part 3 of 3)
745 [2522] localedata: ca_ES@valencia: new Valencian (meridional Catalan)
747 [5997] math: Very slow execution of sinf function
748 [10580] localedata: hr_HR: updated locale
749 [10871] locale: 'mon' array should contain both nominative and genitive
751 [12349] localedata: eu_ES: incorrect thousands separator
752 [13605] localedata: shn_MM: new Shan locale
753 [13805] localedata: ru_RU: currency should use ',' as radix point
754 [13953] localedata: km_KH: locale update
755 [13994] localedata: mjw_IN: new locale
756 [14121] build: make writes .mo files in po directory
757 [14333] libc: Fix the race between atexit() and exit()
758 [14681] dynamic-link: _dl_get_origin leaks memory via executable link map.
759 [14925] localedata: bn_*: LC_IDENTIFICATION.language key should be
761 [15260] localedata: LC_MESSAGES.{yes,no}{str,expr}: various errors
762 [15261] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of full-
763 width Latin characters
764 [15332] localedata: es_CU: locale update
765 [15436] stdio: Don't close or flush stdio streams on abort
766 [15537] localedata: lv_LV: invalid collation for Latvian diacritical
768 [16148] localedata: ca_ES: incorrect thousands separator
769 [16750] dynamic-link: ldd should not try to execute the binaries
771 [16777] localedata: pl_PL: incorrect thousands separator in locale
772 [16905] localedata: hanzi: new collation
773 [17563] localedata: cmn_TW: add hanzi collation
774 [17750] localedata: wrong collation order of diacritics in most locales
775 [17804] libc: scandirat fails with ENOMEM because it checks for errno even
777 [17956] build: Build fails on missing definitions from header file
778 nss/nss.h when Mozilla NSS is used for cryptography
779 [18203] libc: realpath() does not handle unreachable paths correctly
780 [18572] dynamic-link: [arm] Lazy TLSDESC relocation has data race
781 [18812] localedata: kab_DZ: new Kabyle Algeria locale
782 [18822] libc: Internal functions are called via PLT
783 [18858] string: _HAVE_STRING_ARCH_xxx aren't defined for i386 nor x86_64
784 [19170] libc: __gmon_start__ defined in hppa in crtn.S
785 [19574] libc: glibc should support building static PIE binaries
786 [19852] localedata: charmaps/UTF-8: incorrect wcwidth for U+3099 and
788 [19971] glob: glob: Do not skip entries with zero d_ino values
789 [19982] localedata: fr.po: spelling mistake for error code EXDEV
790 [20008] localedata: km_KH: convert to translit_neutral
791 [20009] localedata: tr_TR: convert LC_CTYPE to i18n
792 [20142] math: [x86_64] Add SSE4.1 trunc, truncf
793 [20204] dynamic-link: _dl_open_hook and _dlfcn_hook hardening
794 [20482] localedata: de_CH: abbreviated weekdays should be two letters
795 [20498] localedata: miq_NI: new Mískitu / Miskito (miq) language locale
797 [20532] nss: getaddrinfo uses errno and h_errno without guaranteeing
798 they're set, wrong errors returned by gaih_inet when lookup functions
800 [20756] localedata: [PATCH] Use Unicode wise thousands separator
801 [20826] network: posix/tst-getaddrinfo5 fails on hosts without network
803 [20952] localedata: yuw_PG: new locale
804 [21084] localedata: charmaps/IBM858: new codepage
805 [21161] manual: [PATCH] fix typo in manual/arith.texi on strtoul prototype
806 [21242] libc: assert gives pedantic warning in old gcc versions
807 [21265] dynamic-link: _dl_runtime_resolve isn't compatible with Intel C++
808 __regcall calling convention
809 [21309] math: signed integer overflow in sysdeps/ieee754/dbl-64/e_pow.c
810 [21326] libc: C99 functions are not declared for C++11 and later if
811 _GNU_SOURCE is not predefined by g++
812 [21457] libc: sys/ucontext.h namespace
813 [21530] libc: tmpfile() should be implemented using O_TMPFILE
814 [21660] math: GCC fails to compile a formula with tgmath.h
815 [21672] nptl: sys-libs/glibc on ia64 crashes on thread exit: signal
816 SIGSEGV, Segmentation fault: pthread_create.c:432: __madvise
817 (pd->stackblock, freesize - PTHREAD_STACK_MIN, MADV_DONTNEED);
818 [21684] math: tgmath.h handling of complex integers
819 [21685] math: tgmath.h handling of bit-fields
820 [21686] math: tgmath.h handling of __int128
821 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
823 [21745] libc: [powerpc64le] Extra PLT reference with --enable-stack-
825 [21750] localedata: column width of characters incompatible with classical
827 [21754] malloc: malloc: Perform as little work as possible after heap
828 consistency check failures
829 [21780] libc: hppa: p{read,write}v2 does not set ENOSUP on invalid flag
830 [21790] libc: Missing __memset_zero_constant_len_parameter in libc.so
831 [21791] string: Unused XXX_chk_XXX functions in libc.a
832 [21815] dynamic-link: FAIL: elf/tst-prelink-cmp with GCC is defaulted to
834 [21836] localedata: Removed redundant data (LC_MONETARY) in various Indian
836 [21845] localedata: Added new Locale bho_NP
837 [21853] localedata: Fix abday Which looks same as day in zh_SG
838 [21854] localedata: Added New Locale en_SC
839 [21864] libc: xmalloc.o is compiled with -DMODULE_NAME=libc
840 [21871] dynamic-link: _dl_runtime_resolve_avx_opt is slower than
841 _dl_runtime_resolve_avx_slow
842 [21885] network: getaddrinfo: gethosts does not release resolver context
843 on memory allocation failure
844 [21899] libc: XPG4.2 sigaction namespace
845 [21908] dynamic-link: dynamic linker broke on ia64 (mmap2 consolidation is
847 [21913] libc: static binaries SIGSEGV in __brk when host's gcc is pie-by-
849 [21915] nss: nss_files can return with NSS_STATUS_SUCCESS and a clobbered
850 errno value, causing getaddrinfo to fail
851 [21920] localedata: Fix p_cs_precedes/n_cs_precedes for mt_MT
852 [21922] network: getaddrinfo with AF_INET/AF_INET6 returns EAI_NONAME
853 instead of EAI_NODATA
854 [21928] libc: sys/ptrace.h: remove obsolete temporary development Linux
855 constant PTRACE_SEIZE_DEVEL
856 [21930] math: C-only gcc builtins used in <math.h> isinf
857 [21932] network: Unpaired __resolv_context_get in generic get*_r
859 [21941] math: powerpc: Wrong register constraint for xssqrtqp in sqrtf128
860 [21944] libc: sigval namespace
861 [21951] localedata: Update hanzi collation by stroke
862 [21955] math: Wrong alignment of L(SP_RANGE)/L(SP_INF_0) in
863 sysdeps/x86_64/fpu/e_expf.S
864 [21956] libc: Stack allocation in MIPS syscall impl (ubounded stack
865 allocation in syscall loops)
866 [21959] localedata: Fix Country name for xh_ZA
867 [21960] localedata: Fix abmon for bem_ZM
868 [21966] math: AVX2 mathvec functions use FMA without checking
869 [21967] math: When 512-bit AVX2 wrapper functions in mathvec are used?
870 [21971] localedata: Added New Locale for mfe_MU
871 [21972] libc: assert macro requires operator== (int) for its argument type
872 [21973] math: [sparc] libm missing sqrtl compat symbol
873 [21974] libc: Remove __bb_init_func and __bb_exit_func
874 [21982] string: stratcliff.c: error: assuming signed overflow does not
876 [21986] stdio: __guess_grouping is called incorrectly
877 [21987] math: [sparc32] wrong bits/long-double.h installed
878 [22019] localedata: Wrong placement of monetary symbol in el_GR (negative
880 [22022] localedata: Missing country_name for mni_IN
881 [22023] localedata: Removed redundant data (LC_TIME and LC_MESSAGES) for
883 [22025] locale: iconv: Inconsistency between pointer mangling and NULL
885 [22026] locale: iconv_open: heap overflow on gconv_init failure
886 [22028] math: bits/math-finite.h _MSUF_ expansion namespace
887 [22035] math: [m68k] bits/math-inline.h macro namespace
888 [22038] localedata: Fix abbreviated weeks and months for Somali
889 [22044] localedata: Remove redundant data for Limburgish Language
890 [22050] malloc: Linking with -lmcheck does not hook
891 __malloc_initialize_hook correctly
892 [22051] libc: zero terminator in the middle of glibc's .eh_frame
893 [22052] malloc: malloc failed to compile with GCC 7 and -O3
894 [22070] localedata: charmaps/UTF-8: wcwidth for
895 Prepended_Concatenation_Mark codepoints set to 0 (should be 1)
896 [22074] localedata: charmaps/UTF-8: wcwidth for U+1160-U+11FF (Hangul
897 Jungseong and Jongseong) should be 0
898 [22078] nss: nss_files performance issue in multi mode
899 [22082] math: bits/math-finite.h exp10 condition
900 [22086] libc: pcprofiledump incorrect cross-endian condition
901 [22093] dynamic-link: ld.so no longer searches in .../x86_64
902 [22095] network: Name server address allocation memory leak in resolv.conf
904 [22096] network: __resolv_conf_attach can incorrectly free passed conf
906 [22100] localedata: om_KE: LC_TIME: copy redundant data from om_ET
907 [22101] dynamic-link: Dynamic loader must ignore "debug" shared objects
909 [22111] malloc: malloc: per thread cache is not returned when thread exits
910 [22112] localedata: Fix LC_TELEPHONE/LC_NAME for az_AZ
911 [22134] libc: [linux] implement fexecve with execveat
912 [22142] libc: [powerpc] printf oupts a wrong value of DBL_MAX on ppc64 and
914 [22145] libc: ttyname() gives up too early in the face of namespaces
915 [22146] math: C++ build issue with float128 on x86_64
916 [22153] nptl: nptl: save error code before process termination
917 [22156] libc: [hppa,ia64,microblaze] Executable stack default
918 [22159] malloc: malloc: MALLOC_CHECK_ broken with --enable-tunables=no
919 [22161] nscd: nscd cache prune for netgroups hangs after timeout bump
920 [22165] libc: [hppa] Text relocations in libc.so
921 [22180] libc: destructor registered via __cxa_atexit is called twice
922 [22183] glob: commit 5554304f0ddd ("posix: Allow glob to match dangling
923 symlinks") cause "make" segfaults
924 [22189] math: [powerpc] math_private.h definitions of math_opt_barrier and
926 [22207] libc: FAIL: stdlib/test-atexit-race
927 [22225] math: nearbyint arithmetic moved before feholdexcept
928 [22229] math: [sparc32] missing copysignl, fabsl, fmal compat symbols
929 [22235] math: iscanonical in C++ and float128
930 [22243] math: log2(0) and log10(0) are wrong in downward rounding without
931 the svid compat wrapper
932 [22244] math: ynf and yn are wrong without the svid compat wrapper
933 [22273] libc: Improper assert in Linux posix_spawn implementation
934 [22284] libc: -pg -pie doesn't work
935 [22292] locale: localedef exits with error 4 when it should be error 1
936 [22294] locale: Allow "" for int_currency_symbol definition in locales.
937 [22295] locale: Don't warn on non-symbolic characters in locale sources in
939 [22296] math: glibc 2.26: signbit build issue with Gcc 5.5.0 on x86_64
940 [22298] nptl: x32: lockups on recursive pthread_mutex_lock after upgrade
942 [22299] dynamic-link: Problem with $PLATFORM on x86_64 platform
943 [22320] glob: Buffer overflow in glob with GLOB_TILDE (CVE-2017-15670)
944 [22321] libc: sysconf(_SC_IOV_MAX) returns -1 on Linux
945 [22322] libc: [mips64] wrong bits/long-double.h installed
946 [22325] glob: Memory leak in glob with GLOB_TILDE (CVE-2017-15671)
947 [22332] glob: Buffer overflow in glob with GLOB_TILDE in unescaping
949 [22336] localedata: cs_CZ LC_COLLATE does not use i18n
950 [22343] malloc: Integer overflow in posix_memalign
951 [22347] libc: getrandom() returns the number of bytes that were copied to
952 the buffer even though the comments say "Return 0 on success and -1 on
954 [22353] string: sysdeps/i386/i586/strcpy.S isn't maintainable
955 [22362] libc: Installed crt1.o, crti,.o and crtn.o files are used with
957 [22370] dynamic-link: Incorrect note padding check
958 [22375] libc: malloc returns pointer from tcache_get when should return
959 NULL (CVE-2017-17426)
960 [22377] math: iseqsig, float128 and C++
961 [22382] localedata: Error in tpi_PG locale
962 [22387] localedata: Replace unicode sequences <Uxxxx> for characters
963 inside the ASCII printable range
964 [22402] math: [powerpc64le] __MATH_TG does not support _Float128 for
966 [22403] localedata: Slash needs escaping in some locales
967 [22408] malloc: malloc_info access heaps without arena lock, ignores heaps
968 [22409] network: res_hnok does not accept some host names used on the
970 [22412] network: res_dnok, res_hnok should perform syntax checks
971 [22413] network: ns_name_pton ignores syntactically invalid trailing
973 [22415] stdio: setvbuf can lead to invalid free/segfault
974 [22432] build: Non-deterministic build
975 [22439] malloc: malloc_info should compute summary statistics for all sub-
977 [22442] network: if_nametoindex could report index for the wrong
979 [22446] build: aliasing violation calling readlink in handle_request
980 [22447] build: unsafe call to strlen with a non-string in getlogin_r.c
981 [22457] libc: Generic preadv/pwritev incorrectly calls __posix_memalign
982 [22459] libc: FAIL: elf/check-localplt with __stack_chk_fail related to
983 __nscd_hash/__nss_hash
984 [22463] network: p_secstodate overflow handling
985 [22469] localedata: pl_PL LC_COLLATE does not use i18n
986 [22478] libc: sigwait can fail with EINTR
987 [22505] libc: ldconfig processes include directive in locale-specific
989 [22515] localedata: hsb_DE LC_COLLATE does not use copy "iso14651_t1"
990 [22517] localedata: et_EE LC_COLLATE does not use copy "iso14651_t1"
991 [22519] localedata: is_IS LC_COLLATE does not use copy "iso14651_t1"
992 [22524] localedata: lt_LT LC_COLLATE does not use copy "iso14651_t1"
993 [22527] localedata: tr_TR LC_COLLATE does not use copy "iso14651_t1"
994 [22534] localedata: Collation rules for Serbian and Bosnian should be the
996 [22561] math: [DR#471] cacosh (0 + iNaN) should return NaN +/- i pi/2
997 [22568] math: [DR#471] ctanh (0 + iNaN), ctanh (0 + i Inf)
998 [22577] libc: missing newline after "cannot allocate TLS data structures
1000 [22588] manual: manual/conf.texi: missing underscore in front of
1002 [22593] math: nextafter and nexttoward are declared with const attribute
1003 [22596] manual: manual: finite(nan) wrongly described as returning nonzero
1004 [22603] string: ia64 memchr overflows internal pointer check
1005 [22605] libc: SH clone does not set the exit code correctly
1006 [22606] dynamic-link: Incorrect array size computation in _dl_init_paths
1008 [22607] dynamic-link: Buffer Overflow in _dl_init_paths (CVE-2017-1000409)
1009 [22611] malloc: malloc/tst-realloc wrongly assumes that errno must not be
1010 modified in case of success
1011 [22614] build: gcc: error: unrecognized command line option ‘-no-pie’
1012 [22615] manual: manual: ambiguous wording about errno value in case of
1014 [22624] libc: MIPS setjmp() saves incorrect 'o0' register in --enable-
1016 [22625] dynamic-link: RPATH $ORIGIN replaced by PWD for AT_SECURE/SUID
1017 binaries or if /proc is not mounted (CVE-2017-16997)
1018 [22627] dynamic-link: $ORIGIN in $LD_LIBRARY_PATH is substituted twice
1019 [22630] build: $(no-pie-ldflag) is no longer effective
1020 [22631] math: [m68k] Bad const attributes in bits/mathinline.h
1021 [22635] nptl: pthread_self returns NULL before libpthread is loaded
1022 [22636] nptl: PTHREAD_STACK_MIN is too small on x86-64
1023 [22637] nptl: guard size is subtracted from thread stack size instead of
1025 [22648] libc: getrlimit/setrlimit with RLIM_INFINITY broken on alpha
1026 [22657] localedata: hu_HU: Avoid double space in date
1027 [22660] math: fmax, fmin sNaN handling on alpha
1028 [22664] libc: New warning of GCC8
1029 [22665] math: alpha: ceil and floor raise inexact exceptions
1030 [22666] math: alpha: trunc raise inexact exceptions
1031 [22667] libc: makecontext lacks stack alignment on i386
1032 [22678] libc: prlimit fails for RLIM_INFINITY values on 32-bit machines
1033 [22679] libc: getcwd(3) can succeed without returning an absolute path
1035 [22685] libc: PowerPC: Static AT_SECURE binaries segfault with lock-
1036 elision and tunables
1037 [22687] math: [powerpc-nofpu] complex long double functions spurious
1039 [22688] math: [powerpc-nofpu] remainderl wrong sign of zero result
1040 [22690] math: [ldbl-128ibm] lrintl, lroundl missing "invalid" exceptions
1041 [22691] math: [powerpc-nofpu] fmaxmagl, fminmagl spurious "invalid"
1043 [22693] math: [ldbl-128ibm] log1pl (-qNaN) spurious "invalid" exception
1044 [22697] math: [powerpc] llround spurious "inexact" exceptions on 32-bit
1046 [22701] nis: Incomplete removal of libnsl
1047 [22702] math: [powerpc-nofpu] nearbyintl traps with trapping "inexact"
1048 [22707] libc: Missing defines in elf.h for DF_1_STUB and DF_1_PIE.
1049 [22715] dynamic-link: FAIL: elf/tst-audit10
1050 [22719] libc: Backtrace tests fail on hppa
1051 [22742] libc: [aarch64] mcontext_t __reserved field got renamed
1052 [22743] nptl: __pthread_register_cancel corrupts stack after f81ddabffd
1053 [22765] crypt: (struct crypt_data *data)->initialized is not set to zero
1054 before the first call to crypt_r () in crypt/badsalttest.c
1061 * A per-thread cache has been added to malloc. Access to the cache requires
1062 no locks and therefore significantly accelerates the fast path to allocate
1063 and free small amounts of memory. Refilling an empty cache requires locking
1064 the underlying arena. Performance measurements show significant gains in a
1065 wide variety of user workloads. Workloads were captured using a special
1066 instrumented malloc and analyzed with a malloc simulator. Contributed by
1067 DJ Delorie with the help of Florian Weimer, and Carlos O'Donell.
1069 * Unicode 10.0.0 Support: Character encoding, character type info, and
1070 transliteration tables are all updated to Unicode 10.0.0, using
1071 generator scripts contributed by Mike FABIAN (Red Hat).
1072 These updates cause user visible changes, especially the changes in
1073 wcwidth for many emoji characters cause problems when emoji sequences
1074 are rendered with pango, see for example:
1075 https://bugzilla.gnome.org/show_bug.cgi?id=780669#c5
1077 * Collation of Hungarian has been overhauled and is now consistent with "The
1078 Rules of Hungarian Orthography, 12th edition" (Bug 18934). Contributed by
1081 * Improvements to the DNS stub resolver, contributed by Florian Weimer:
1083 - The GNU C Library will now detect when /etc/resolv.conf has been
1084 modified and reload the changed configuration. The new resolver option
1085 “no-reload” (RES_NORELOAD) disables this behavior.
1087 - The GNU C Library now supports an arbitrary number of search domains
1088 (configured using the “search” directive in /etc/resolv.conf);
1089 previously, there was a hard limit of six domains. For backward
1090 compatibility, applications that directly modify the ‘_res’ global
1091 object are still limited to six search domains.
1093 - When the “rotate” (RES_ROTATE) resolver option is active, the GNU C
1094 Library will now randomly pick a name server from the configuration as a
1095 starting point. (Previously, the second name server was always used.)
1097 * The tunables feature is now enabled by default. This allows users to tweak
1098 behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
1100 * New function reallocarray, which resizes an allocated block (like realloc)
1101 to the product of two sizes, with a guaranteed clean failure upon integer
1102 overflow in the multiplication. Originally from OpenBSD, contributed by
1103 Dennis Wölfing and Rüdiger Sonderfeld.
1105 * New wrappers for the Linux-specific system calls preadv2 and pwritev2.
1106 These are extended versions of preadv and pwritev, respectively, taking an
1107 additional flags argument. The set of supported flags depends on the
1108 running kernel; full support currently requires kernel 4.7 or later.
1110 * posix_spawnattr_setflags now supports the flag POSIX_SPAWN_SETSID, to
1111 create a new session ID for the spawned process. This feature is
1112 scheduled to be added to the next major revision of POSIX; for the time
1113 being, it is available under _GNU_SOURCE.
1115 * errno.h is now safe to use from C-preprocessed assembly language on all
1116 supported operating systems. In this context, it will only define the
1117 Exxxx constants, as preprocessor macros expanding to integer literals.
1119 * On ia64, powerpc64le, x86-32, and x86-64, the math library now implements
1120 128-bit floating point as defined by ISO/IEC/IEEE 60559:2011 (IEEE
1121 754-2008) and ISO/IEC TS 18661-3:2015. Contributed by Paul E. Murphy,
1122 Gabriel F. T. Gomes, Tulio Magno Quites Machado Filho, and Joseph Myers.
1124 To compile programs that use this feature, the compiler must support
1125 128-bit floating point with the type name _Float128 (as defined by TS
1126 18661-3) or __float128 (the nonstandard name used by GCC for C++, and for
1127 C prior to version 7). _GNU_SOURCE or __STDC_WANT_IEC_60559_TYPES_EXT__
1128 must be defined to make the new interfaces visible.
1130 The new functions and macros correspond to those present for other
1131 floating-point types (except for a few obsolescent interfaces not
1132 supported for the new type), with F128 or f128 suffixes; for example,
1133 strtof128, HUGE_VAL_F128 and cosf128. Following TS 18661-3, there are no
1134 printf or scanf formats for the new type; the strfromf128 and strtof128
1135 interfaces should be used instead.
1137 Deprecated and removed features, and other changes affecting compatibility:
1139 * The synchronization that pthread_spin_unlock performs has been changed to
1140 now be equivalent to a C11 atomic store with release memory order to the
1141 spin lock's memory location. Previously, several (but not all)
1142 architectures used stronger synchronization (e.g., containing what is
1143 often called a full barrier). This change can improve performance, but
1144 may affect odd fringe uses of spin locks that depend on the previous
1145 behavior (e.g., using spin locks as atomic variables to try to implement
1146 Dekker's mutual exclusion algorithm).
1148 * The port to Native Client running on ARMv7-A (--host=arm-nacl) has been
1151 * Sun RPC is deprecated. The rpcgen program, librpcsvc, and Sun RPC headers
1152 will only be built and installed when the GNU C Library is configured with
1153 --enable-obsolete-rpc. This allows alternative RPC implementations, such
1154 as TIRPC or rpcsvc-proto, to be used.
1156 * The NIS(+) name service modules, libnss_nis, libnss_nisplus, and
1157 libnss_compat, are deprecated, and will not be built or installed by
1160 The NIS(+) support library, libnsl, is also deprecated. By default, a
1161 compatibility shared library will be built and installed, but not headers
1162 or development libraries. Only a few NIS-related programs require this
1163 library. (In particular, the GNU C Library has never required programs
1164 that use 'gethostbyname' to be linked with libnsl.)
1166 Replacement implementations based on TIRPC, which additionally support
1167 IPv6, are available from <https://github.com/thkukuk/>. The configure
1168 option --enable-obsolete-nsl will cause libnsl's headers, and the NIS(+)
1169 name service modules, to be built and installed.
1171 * The DNS stub resolver no longer performs EDNS fallback. If EDNS or DNSSEC
1172 support is enabled, the configured recursive resolver must support EDNS.
1173 (Responding to EDNS-enabled queries with responses which are not
1174 EDNS-enabled is fine, but FORMERR responses are not.)
1176 * res_mkquery and res_nmkquery no longer support the IQUERY opcode. DNS
1177 servers have not supported this opcode for a long time.
1179 * The _res_opcodes variable has been removed from libresolv. It had been
1180 exported by accident.
1182 * <string.h> no longer includes inline versions of any string functions,
1183 as this kind of optimization is better done by the compiler. The macros
1184 __USE_STRING_INLINES and __NO_STRING_INLINES no longer have any effect.
1186 * The nonstandard header <xlocale.h> has been removed. Most programs should
1187 use <locale.h> instead. If you have a specific need for the definition of
1188 locale_t with no other declarations, please contact
1189 libc-alpha@sourceware.org and explain.
1191 * The obsolete header <sys/ultrasound.h> has been removed.
1193 * The obsolete signal constant SIGUNUSED is no longer defined by <signal.h>.
1195 * The obsolete function cfree has been removed. Applications should use
1198 * The stack_t type no longer has the name struct sigaltstack. This changes
1199 the C++ name mangling for interfaces involving this type.
1201 * The ucontext_t type no longer has the name struct ucontext. This changes
1202 the C++ name mangling for interfaces involving this type.
1204 * On M68k GNU/Linux and MIPS GNU/Linux, the fpregset_t type no longer has
1205 the name struct fpregset. On Nios II GNU/Linux, the mcontext_t type no
1206 longer has the name struct mcontext. On SPARC GNU/Linux, the struct
1207 mc_fq, struct rwindow, struct fpq and struct fq types are no longer
1208 defined in sys/ucontext.h, the mc_fpu_t type no longer has the name struct
1209 mc_fpu, the gwindows_t type no longer has the name struct gwindows and the
1210 fpregset_t type no longer has the name struct fpu. This changes the C++
1211 name mangling for interfaces involving those types.
1213 * On S/390 GNU/Linux, the constants defined by <sys/ptrace.h> have been
1214 synced with the kernel:
1216 - PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS
1217 are not supported on this architecture and have been removed.
1219 - PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
1220 PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
1221 PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND have been added.
1223 Programs that assume the GET/SETREGS ptrace requests are universally
1224 available will now fail to build, instead of malfunctioning at runtime.
1226 Changes to build and runtime requirements:
1228 * Linux kernel 3.2 or later is required at runtime, on all architectures
1229 supported by that kernel. (This is a change from version 2.25 only for
1232 * GNU Binutils 2.25 or later is now required to build the GNU C Library.
1234 * On most architectures, GCC 4.9 or later is required to build the GNU C
1235 Library. On powerpc64le, GCC 6.2 or later is required.
1237 Older GCC versions and non-GNU compilers are still supported when
1238 compiling programs that use the GNU C Library. (We do not know exactly
1239 how old, and some GNU extensions to C may be _de facto_ required. If you
1240 are interested in helping us make this statement less vague, please
1241 contact libc-alpha@sourceware.org.)
1243 Security related changes:
1245 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
1246 to avoid fragmentation-based spoofing attacks (CVE-2017-12132).
1248 * LD_LIBRARY_PATH is now ignored in binaries running in privileged AT_SECURE
1249 mode to guard against local privilege escalation attacks (CVE-2017-1000366).
1251 * Avoid printing a backtrace from the __stack_chk_fail function since it is
1252 called on a corrupt stack and a backtrace is unreliable on a corrupt stack
1255 * A use-after-free vulnerability in clntudp_call in the Sun RPC system has been
1256 fixed (CVE-2017-12133).
1258 The following bugs are resolved with this release:
1260 [984] network: Respond to changed resolv.conf in gethostbyname
1261 [5010] network: sunrpc service cleanup causes unwanted port mapper traffic
1262 [12068] localedata: sc_IT: misspelled yesexpr/day/abday/mon/abmon/date_fmt
1264 [12189] libc: __stack_chk_fail should not attempt a backtrace
1266 [14096] time: Race condition on timezone/tst-timezone.out
1267 [14172] localedata: az_IR: new locale
1268 [14995] build: glibc fails to build if gold is the default linker, even if
1270 [15998] build: [powerpc] Set arch_minimum_kernel for powerpc LE
1271 [16637] network: inet_pton function is accepting IPv6 with bad format
1272 [16640] string: string/strtok.c: undefined behaviour inconsistent between
1273 x86 and other generic code
1274 [16875] localedata: ko_KR: fix lang_name
1275 [17225] localedata: ar_SY: localized month names for May and June are
1277 [17297] localedata: da_DK: wrong date_fmt string
1278 [18907] stdio: Incorrect order of __wur __THROW in <printf.h>
1279 [18934] localedata: hu_HU: collate: fix multiple bugs and add tests
1280 [18988] nptl: pthread wastes memory with mlockall(MCL_FUTURE)
1281 [19066] localedata: ar_SA abbreviated day and month names are in English
1282 [19569] network: resolv: Support an arbitrary number of search domains
1283 [19570] network: Implement random DNS server selection in the stub
1285 [19838] locale: localedef fails on PA-RISC
1286 [19919] localedata: iso14651_t1_common: Correct the Malayalam sorting
1287 order of 0D36 and 0D37
1288 [19922] localedata: iso14651_t1_common: Define collation for Malayalam
1290 [20098] libc: FAIL: debug/backtrace-tst on hppa
1291 [20257] network: sunrpc: clntudp_call does not enforce timeout when
1293 [20275] localedata: locale day/abday/mon/abmon should not have trailing
1295 [20313] localedata: Update locale data to Unicode 9.0
1296 [20424] manual: Document how to provide a malloc replacement
1297 [20496] localedata: agr_PE: new language locale Awajún / Aguaruna (agr)
1299 [20686] locale: Add el_GR@euro to SUPPORTED.
1300 [20831] dynamic-link: _dl_map_segments does not test for __mprotect
1301 failures consistently
1302 [21015] dynamic-link: Document and fix --enable-bind-now
1303 [21016] nptl: pthread_cond support is broken on hppa
1304 [21029] libc: glibc-2.23 (and later) fails to compile with -fno-omit-
1305 frame-pointer on i386
1306 [21049] libc: segfault in longjmp_chk() due to clobbered processor
1308 [21075] libc: unused assigment to %g4 in sparc/sparc{64,32}/clone.S
1309 [21088] libc: Build fails with --enable-static-nss
1310 [21094] math: cosf(1.57079697) has 3 ulp error on targets where the
1311 generic c code is used
1312 [21109] libc: Tunables broken on big-endian
1313 [21112] math: powf has large ulp errors with base close to 1 and exponent
1315 [21115] network: sunrpc: Use-after-free in error path in clntudp_call
1317 [21120] malloc: glibc malloc is incompatible with GCC 7
1318 [21130] math: Incorrect return from y0l (-inf) and y1l (-inf) when linking
1320 [21134] math: Exception (divide by zero) not set for y0/y1 (0.0) and y0/y1
1321 (-0.0) when linking with -lieee
1322 [21171] math: log10, log2 and lgamma return incorrect results
1323 [21179] libc: handle R_SPARC_DISP64 and R_SPARC_REGISTER relocs
1324 [21182] libc: __memchr_sse2: regression in glibc-2.25 on i686
1325 [21207] localedata: ce_RU: update weekdays from CLDR
1326 [21209] dynamic-link: LD_HWCAP_MASK read in setuid binaries
1327 [21217] localedata: Update months from CLDR-31
1328 [21232] libc: miss posix_fadvise64 on MIPS64 when static linking
1329 [21243] libc: support_delete_temp_file should issue warning for failed
1331 [21244] libc: support resolv_test_start() socket fd close should be
1333 [21253] libc: localedef randomly segfaults when using -fstack-check due to
1334 new posix_spawn implementation
1335 [21258] dynamic-link: Branch predication in _dl_runtime_resolve_avx512_opt
1336 leads to lower CPU frequency
1337 [21259] libc: [alpha] termios.h missing IXANY for POSIX
1338 [21261] libc: [sparc64] bits/setjmp.h namespace
1339 [21267] network: [mips] bits/socket.h IOC* namespace
1340 [21268] libc: [alpha] termios.h NL2, NL3 namespace
1341 [21270] libc: mmap64 silently truncates large offset values
1342 [21275] libc: posix_spawn always crashes on ia64 now
1343 [21277] libc: [alpha] termios.h missing IUCLC for UNIX98 and older
1344 [21280] math: [powerpc] logbl for POWER7 return incorrect results
1345 [21289] libc: Incorrect declaration for 32-bit platforms with
1346 _FILE_OFFSET_BITS=64 causes build error
1347 [21295] network: GETAI(AF_UNSPEC) drops IPv6 addresses if nss module does
1348 not support gethostbyname4_r
1349 [21298] nptl: rwlock can deadlock on frequent reader/writer phase
1351 [21338] malloc: mallopt M_ARENA_MAX doesn't set the maximum number of
1353 [21340] libc: Support POSIX_SPAWN_SETSID
1354 [21357] libc: unwind-dw2-fde deadlock when using AddressSanitizer
1355 [21359] network: ns_name_pack needs additional byte in destination buffer
1356 [21361] network: resolv: Reduce advertised EDNS0 buffer size to guard
1357 against fragmentation attacks (CVE-2017-12132)
1358 [21369] network: resolv: Remove EDNS fallback
1359 [21371] libc: Missing timespec definition when compiled with _XOPEN_SOURCE
1361 [21386] nptl: Assertion in fork for distinct parent PID is incorrect
1362 [21391] dynamic-link: x86: Set dl_platform and dl_hwcap from CPU features
1363 [21393] stdio: Missing dup3 error check in freopen, freopen64
1364 [21396] libc: Use AVX2 memcpy/memset on Skylake server
1365 [21399] localedata: Bad description for U00EC in
1366 localedata/charmaps/CP1254
1367 [21411] malloc: realloc documentation error
1368 [21426] network: sys/socket.h uio.h namespace
1369 [21428] libc: [aarch64] tst-backtrace5 testsuite failure
1370 [21445] libc: signal.h bsd_signal namespace
1371 [21455] network: Network headers stdint.h namespace
1372 [21474] network: resolv: res_init does not use RES_DFLRETRY (2) but 4 for
1374 [21475] network: resolv: Overlong search path is truncated mid-label
1375 [21511] libc: sigstack namespace
1376 [21512] libc: clone() ends up calling exit_group() through _exit() wrapper
1377 [21514] libc: sysdeps/unix/sysv/linux/sys/syscall.h:31:27: fatal error:
1378 bits/syscall.h: No such file or directory
1379 [21517] libc: struct sigaltstack namespace
1380 [21528] dynamic-link: Duplicated minimal strtoul implementations in ld.so
1381 [21533] localedata: Update locale data to Unicode 10.0
1383 ../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44: Error:
1384 junk at end of line, first unrecognized character is `@'
1385 [21538] libc: SIG_HOLD missing for XPG4
1386 [21539] libc: S390: Mismatch between kernel and glibc ptrace.h with
1387 request 12: PTRACE_SINGLEBLOCK vs PTRACE_GETREGS.
1388 [21542] libc: Use conservative default for sysconf (_SC_NPROCESSORS_ONLN)
1389 [21543] libc: sigevent namespace
1390 [21548] libc: [mips] get/set/make/swap context for MIPS O32 assume wrong
1391 size for general purpose registers in mcontext_t structure
1392 [21550] libc: sigwait namespace
1393 [21552] libc: XPG4 bsd_signal namespace
1394 [21554] libc: sigpause namespace
1395 [21560] libc: sys/wait.h signal.h namespace
1396 [21561] libc: waitid namespace
1397 [21573] nptl: GCC 7: /usr/bin/install: cannot remove
1398 '/usr/include/stdlib.h': Permission denied
1399 [21575] libc: sys/wait.h missing struct rusage definition
1400 [21584] libc: sigaltstack etc namespace
1401 [21597] libc: siginterrupt namespace
1402 [21607] math: hppa: FAIL: math/test-tgmath
1403 [21609] dynamic-link: Incomplete workaround for GCC __tls_get_addr ABI
1405 [21622] libc: [tile] missing SA_* for POSIX.1:2008
1406 [21624] dynamic-link: ld.so: Unsafe alloca allows local attackers to alias
1407 stack and heap (CVE-2017-1000366)
1408 [21625] libc: wait3 namespace
1409 [21654] nss: Incorrect pointer alignment in NSS group merge result
1411 [21657] network: Parse interface zone id for node-local multicast
1412 [21662] string: memcmp-avx2-movbe.S lacks saturating subtraction for
1414 [21666] libc: .symver is used on common symbol
1415 [21668] network: resolv: res_init cross-thread broadcast introduces race
1417 [21687] math: tgmath.h totalorder, totalordermag return type
1418 [21694] locale: Current Glibc Locale Does Not Support Tok-Pisin and Fiji
1420 [21696] libc: Incorrect assumption of of __cpu_mask in
1421 posix/sched_cpucount.c
1422 [21697] libc: sysdeps/posix/spawni.c: 2 * suspicious condition ?
1423 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1425 [21707] math: ppc64le: Invalid IFUNC resolver from libgcc calls getauxval,
1426 leading to relocation crash
1427 [21709] libc: resolv_conf.c:552: update_from_conf: Assertion
1428 `resolv_conf_matches (resp, conf)' failed.
1429 [21710] localedata: Added Samoan language locale for Samoa
1430 [21711] localedata: Pashto yesstr/nostr locale are missing
1431 [21715] nptl: sysdeps/nptl/bits/pthreadtypes.h: typedef guard
1432 __have_pthread_attr_t can cause redefinition of typedef ‘pthread_attr_t’
1433 [21721] localedata: Incorrect Full Weekday names for ks_IN@devanagari
1434 [21723] localedata: yesstr/nostr missing for Chinese language locale
1435 [21724] localedata: yesstr and nostr are missing for Xhosa [LC_MESSAGES]
1437 [21727] localedata: yesstr and nostr are missing for Tsonga [LC_MESSAGES]
1439 [21728] localedata: New Locale for Tongan language
1440 [21729] localedata: incorrect LC_NAME fields for hi_IN
1441 [21733] localedata: yesstr and nostr are missing for zh_HK
1442 [21734] localedata: Missing yesstr and nostr are for kw_GB
1443 [21738] libc: misc/tst-preadvwritev2 and misc/tst-preadvwritev64v2 fail
1444 [21741] libc: Undefined __memmove_chk_XXX and __memset_chk_XXX in libc.a
1445 [21742] libc: _dl_num_cache_relocations is undefined in libc.a
1446 [21743] localedata: ks_IN@devanagari: abday strings mismatch the day
1448 [21744] libc: Tests failing on --enable-tunables --enable-stack-
1450 [21749] localedata: Wrong abbreviated day name (“abday”) for
1452 [21756] localedata: missing yesstr, nostr for nds_DE and nds_NL
1453 [21757] localedata: missing yesstr, nostr for pap_AW and pap_CW
1454 [21759] localedata: missing yesstr and nostr for Tigrinya
1455 [21760] localedata: Fix LC_MESSAGES and LC_ADDRESS for anp_IN
1456 [21766] localedata: Wrong LC_MESSAGES for om_ET Locale
1457 [21767] localedata: Missing Bislama locales
1458 [21768] localedata: Missing yesstr and nostr for aa_ET
1459 [21770] localedata: Missing Field in li_NL
1460 [21778] nptl: Robust mutex may deadlock
1461 [21779] libc: MicroBlaze segfaults when loading libpthread
1462 [21783] localedata: Fix int_select international_call_prefixes
1463 [21784] localedata: Inconsistency in country_isbn
1464 [21788] localedata: Missing Country Postal Abbreviations
1465 [21794] localedata: Added-country_isbn-for-Italy
1466 [21795] localedata: Add/Fix country_isbn for France
1467 [21796] localedata: Added country_isbn for Republic of Korea
1468 [21797] localedata: Fix inconsistency in country_isbn and missing prefixes
1469 [21799] localedata: Added int_select international_call_prefixes
1470 [21801] localedata: Added int_select international_call_prefixes
1471 [21804] nptl: Double semicolon in thread-shared-types.h
1472 [21807] localedata: LC_ADDRESS fix for pap_CW
1473 [21808] localedata: Fix LC_ADDRESS for pap_AW
1474 [21821] localedata: Added country_name in mai_IN
1475 [21822] localedata: Fix LC_TIME for mai_IN
1476 [21823] localedata: missing yesstr, nostr for sa_IN
1477 [21825] localedata: Fix name_mrs for mag_IN
1478 [21828] localedata: 2.26 changelog should mention user visible changes
1480 [21835] localedata: Added Maithili language locale for Nepal
1481 [21838] localedata: Removed redundant data for the_NP
1482 [21839] localedata: Fix LC_MONETARY for ta_LK
1483 [21844] localedata: Fix Latin characters and Months Sequence.
1484 [21848] localedata: Fix mai_NP Title Name
1489 * The feature test macro __STDC_WANT_LIB_EXT2__, from ISO/IEC TR
1490 24731-2:2010, is supported to enable declarations of functions from that
1491 TR. Note that not all functions from that TR are supported by the GNU C
1494 * The feature test macro __STDC_WANT_IEC_60559_BFP_EXT__, from ISO/IEC TS
1495 18661-1:2014, is supported to enable declarations of functions and macros
1496 from that TS. Note that not all features from that TS are supported by
1499 * The feature test macro __STDC_WANT_IEC_60559_FUNCS_EXT__, from ISO/IEC TS
1500 18661-4:2015, is supported to enable declarations of functions and macros
1501 from that TS. Note that most features from that TS are not supported by
1504 * The nonstandard feature selection macros _REENTRANT and _THREAD_SAFE are
1505 now treated as compatibility synonyms for _POSIX_C_SOURCE=199506L.
1506 Since the GNU C Library defaults to a much newer revision of POSIX, this
1507 will only affect programs that specifically request an old conformance
1508 mode. For instance, a program compiled with -std=c89 -D_REENTRANT will
1509 see a change in the visible declarations, but a program compiled with
1510 just -D_REENTRANT, or -std=c99 -D_POSIX_C_SOURCE=200809L -D_REENTRANT,
1513 Some C libraries once required _REENTRANT and/or _THREAD_SAFE to be
1514 defined by all multithreaded code, but glibc has not required this for
1517 * The inclusion of <sys/sysmacros.h> by <sys/types.h> is deprecated. This
1518 means that in a future release, the macros “major”, “minor”, and “makedev”
1519 will only be available from <sys/sysmacros.h>.
1521 These macros are not part of POSIX nor XSI, and their names frequently
1522 collide with user code; see for instance glibc bug 19239 and Red Hat bug
1523 130601. <stdlib.h> includes <sys/types.h> under _GNU_SOURCE, and C++ code
1524 presently cannot avoid being compiled under _GNU_SOURCE, exacerbating the
1527 * New <fenv.h> features from TS 18661-1:2014 are added to libm: the
1528 fesetexcept, fetestexceptflag, fegetmode and fesetmode functions, the
1529 femode_t type and the FE_DFL_MODE and FE_SNANS_ALWAYS_SIGNAL macros.
1531 * Integer width macros from TS 18661-1:2014 are added to <limits.h>:
1532 CHAR_WIDTH, SCHAR_WIDTH, UCHAR_WIDTH, SHRT_WIDTH, USHRT_WIDTH, INT_WIDTH,
1533 UINT_WIDTH, LONG_WIDTH, ULONG_WIDTH, LLONG_WIDTH, ULLONG_WIDTH; and to
1534 <stdint.h>: INT8_WIDTH, UINT8_WIDTH, INT16_WIDTH, UINT16_WIDTH,
1535 INT32_WIDTH, UINT32_WIDTH, INT64_WIDTH, UINT64_WIDTH, INT_LEAST8_WIDTH,
1536 UINT_LEAST8_WIDTH, INT_LEAST16_WIDTH, UINT_LEAST16_WIDTH,
1537 INT_LEAST32_WIDTH, UINT_LEAST32_WIDTH, INT_LEAST64_WIDTH,
1538 UINT_LEAST64_WIDTH, INT_FAST8_WIDTH, UINT_FAST8_WIDTH, INT_FAST16_WIDTH,
1539 UINT_FAST16_WIDTH, INT_FAST32_WIDTH, UINT_FAST32_WIDTH, INT_FAST64_WIDTH,
1540 UINT_FAST64_WIDTH, INTPTR_WIDTH, UINTPTR_WIDTH, INTMAX_WIDTH,
1541 UINTMAX_WIDTH, PTRDIFF_WIDTH, SIG_ATOMIC_WIDTH, SIZE_WIDTH, WCHAR_WIDTH,
1544 * New <math.h> features are added from TS 18661-1:2014:
1546 - Signaling NaN macros: SNANF, SNAN, SNANL.
1548 - Nearest integer functions: roundeven, roundevenf, roundevenl, fromfp,
1549 fromfpf, fromfpl, ufromfp, ufromfpf, ufromfpl, fromfpx, fromfpxf,
1550 fromfpxl, ufromfpx, ufromfpxf, ufromfpxl.
1552 - llogb functions: the llogb, llogbf and llogbl functions, and the
1553 FP_LLOGB0 and FP_LLOGBNAN macros.
1555 - Max-min magnitude functions: fmaxmag, fmaxmagf, fmaxmagl, fminmag,
1558 - Comparison macros: iseqsig.
1560 - Classification macros: iscanonical, issubnormal, iszero.
1562 - Total order functions: totalorder, totalorderf, totalorderl,
1563 totalordermag, totalordermagf, totalordermagl.
1565 - Canonicalize functions: canonicalize, canonicalizef, canonicalizel.
1567 - NaN functions: getpayload, getpayloadf, getpayloadl, setpayload,
1568 setpayloadf, setpayloadl, setpayloadsig, setpayloadsigf, setpayloadsigl.
1570 * The functions strfromd, strfromf, and strfroml, from ISO/IEC TS 18661-1:2014,
1571 are added to libc. They convert a floating-point number into string.
1573 * Most of glibc can now be built with the stack smashing protector enabled.
1574 It is recommended to build glibc with --enable-stack-protector=strong.
1575 Implemented by Nick Alcock (Oracle).
1577 * The function explicit_bzero, from OpenBSD, has been added to libc. It is
1578 intended to be used instead of memset() to erase sensitive data after use;
1579 the compiler will not optimize out calls to explicit_bzero even if they
1580 are "unnecessary" (in the sense that no _correct_ program can observe the
1581 effects of the memory clear).
1583 * On ColdFire, MicroBlaze, Nios II and SH3, the float_t type is now defined
1584 to float instead of double. This does not affect the ABI of any libraries
1585 that are part of the GNU C Library, but may affect the ABI of other
1586 libraries that use this type in their interfaces.
1588 * On x86_64, when compiling with -mfpmath=387 or -mfpmath=sse+387, the
1589 float_t and double_t types are now defined to long double instead of float
1590 and double. These options are not the default, and this does not affect
1591 the ABI of any libraries that are part of the GNU C Library, but it may
1592 affect the ABI of other libraries that use this type in their interfaces,
1593 if they are compiled or used with those options.
1595 * The getentropy and getrandom functions, and the <sys/random.h> header file
1598 * The buffer size for byte-oriented stdio streams is now limited to 8192
1599 bytes by default. Previously, on Linux, the default buffer size on most
1600 file systems was 4096 bytes (and thus remains unchanged), except on
1601 network file systems, where the buffer size was unpredictable and could be
1602 as large as several megabytes.
1604 * The <sys/quota.h> header now includes the <linux/quota.h> header. Support
1605 for the Linux quota interface which predates kernel version 2.4.22 has
1608 * The malloc_get_state and malloc_set_state functions have been removed.
1609 Already-existing binaries that dynamically link to these functions will
1610 get a hidden implementation in which malloc_get_state is a stub. As far
1611 as we know, these functions are used only by GNU Emacs and this change
1612 will not adversely affect already-built Emacs executables. Any undumped
1613 Emacs executables, which normally exist only during an Emacs build, should
1614 be rebuilt by re-running “./configure; make” in the Emacs build tree.
1616 * The “ip6-dotint” and “no-ip6-dotint” resolver options, and the
1617 corresponding RES_NOIP6DOTINT flag from <resolv.h> have been removed.
1618 “no-ip6-dotint” had already been the default, and support for the
1619 “ip6-dotint” option was removed from the Internet in 2006.
1621 * The "ip6-bytestring" resolver option and the corresponding RES_USEBSTRING
1622 flag from <resolv.h> have been removed. The option relied on a
1623 backwards-incompatible DNS extension which was never deployed on the
1626 * The flags RES_AAONLY, RES_PRIMARY, RES_NOCHECKNAME, RES_KEEPTSIG,
1627 RES_BLAST defined in the <resolv.h> header file have been deprecated.
1628 They were already unimplemented.
1630 * The "inet6" option in /etc/resolv.conf and the RES_USE_INET6 flag for
1631 _res.flags are deprecated. The flag was standardized in RFC 2133, but
1632 removed again from the IETF name lookup interface specification in RFC
1633 2553. Applications should use getaddrinfo instead.
1635 * DNSSEC-related declarations and definitions have been removed from the
1636 <arpa/nameser.h> header file, and libresolv will no longer attempt to
1637 decode the data part of DNSSEC record types. Previous versions of glibc
1638 only implemented minimal support for the previous version of DNSSEC, which
1639 is incompatible with the currently deployed version.
1641 * The resource record type classification macros ns_t_qt_p, ns_t_mrr_p,
1642 ns_t_rr_p, ns_t_udp_p, ns_t_xfr_p have been removed from the
1643 <arpa/nameser.h> header file because the distinction between RR types and
1644 meta-RR types is not officially standardized, subject to revision, and
1645 thus not suitable for encoding in a macro.
1647 * The types res_sendhookact, res_send_qhook, re_send_rhook, and the qhook
1648 and rhook members of the res_state type in <resolv.h> have been removed.
1649 The glibc stub resolver did not support these hooks, but the header file
1650 did not reflect that.
1652 * For multi-arch support it is recommended to use a GCC which has
1653 been built with support for GNU indirect functions. This ensures
1654 that correct debugging information is generated for functions
1655 selected by IFUNC resolvers. This support can either be enabled by
1656 configuring GCC with '--enable-gnu-indirect-function', or by
1657 enabling it by default by setting 'default_gnu_indirect_function'
1658 variable for a particular architecture in the GCC source file
1661 * GDB pretty printers have been added for mutex and condition variable
1662 structures in POSIX Threads. When installed and loaded in gdb these pretty
1663 printers show various pthread variables in human-readable form when read
1664 using the 'print' or 'display' commands in gdb.
1666 * Tunables feature added to allow tweaking of the runtime for an application
1667 program. This feature can be enabled with the '--enable-tunables' configure
1668 flag. The GNU C Library manual has details on usage and README.tunables has
1669 instructions on adding new tunables to the library.
1671 * A new version of condition variables functions have been implemented in
1672 the NPTL implementation of POSIX Threads to provide stronger ordering
1675 * A new version of pthread_rwlock functions have been implemented to use a more
1676 scalable algorithm primarily through not using a critical section anymore to
1679 Security related changes:
1681 * On ARM EABI (32-bit), generating a backtrace for execution contexts which
1682 have been created with makecontext could fail to terminate due to a
1683 missing .cantunwind annotation. This has been observed to lead to a hang
1684 (denial of service) in some Go applications compiled with gccgo. Reported
1685 by Andreas Schwab. (CVE-2016-6323)
1687 * The DNS stub resolver functions would crash due to a NULL pointer
1688 dereference when processing a query with a valid DNS question type which
1689 was used internally in the implementation. The stub resolver now uses a
1690 question type which is outside the range of valid question type values.
1693 The following bugs are resolved with this release:
1695 [4099] stdio: Overly agressive caching by stream i/o functions.
1696 [7065] build: Support building glibc with -fstack-protector or -fstack-
1698 [9842] localedata: en_CA: incorrect date format
1699 [13165] nptl: pthread_cond_wait() can consume a signal that was sent
1700 before it started waiting
1701 [14139] manual: Do not hardcode platform names in manual/libm-err-tab.pl
1702 [15765] nptl: sem_open is wrongly a cancellation point
1703 [16421] network: IN6_IS_ADDR_UNSPECIFIED can use undefined s6_addr32
1704 [16458] libc: endian.h and netinet/in.h byte order macros return values of
1706 [16628] dynamic-link: Segfault after a binary without pthread dlopen()s a
1707 library linked with pthread
1708 [16630] nptl: Use SYSENTER for pthread_cond_broadcast/signal() (i.e. fix
1709 "FIXME: Ingo" issue)
1710 [16907] libc: <argp.h> compiled with --std=cXX disables __attribute__
1711 [17252] libc: getrandom and getentropy syscall
1712 [17730] malloc: thread-local storage is sometimes improperly free()'d
1713 after being __libc_memalign()'d
1714 [18241] stdio: failed fseek on memstream does not set errno and can fail
1716 [18243] nptl: sem_wait, sem_timedwait are cancellation points shm_open is
1718 [18463] nptl: pthread_cond_broadcast issue when surrounded by
1719 PTHREAD_PRIO_INHERIT mutex on ARM
1720 [18784] network: res_query and related function crash for special record
1721 type queries (CVE-2015-5180)
1722 [19380] math: strtod does not raise "inexact"
1723 [19387] string: Integer overflow in memchr
1724 [19390] string: Integer overflow in strncat
1725 [19398] build: linknamespace tests fail with massively parallel build
1726 [19402] nptl: Deadlock with robust shared mutex and asynchronous
1728 [19469] malloc: M_PERTURB in test-skeleton.c invalidates malloc tests
1729 [19473] malloc: Turn malloc_get_state etc. in compatibility symbols
1730 [19514] libc: [PATCH] Fix spelling errors in spelling
1731 "implement"/"implementation" in several places
1732 [19582] network: Deprecate RES_USE_INET6
1733 [19673] manual: clog10 docs appear to be erroneous
1734 [19810] dynamic-link: dlopen with both RTLD_NOLOAD and RTLD_NODELETE
1735 causes a segmentation fault
1736 [19826] libc: invalid pointer returned from __tls_get_addr with static
1738 [20016] network: resolv: Remove hooks support from the API
1739 [20019] dynamic-link: NULL pointer dereference in libc.so.6 IFUNC due to
1741 [20033] math: [x86_64] vectorized math function don't call the __finite
1743 [20116] nptl: use after free in pthread_create
1744 [20181] stdio: open_memstream(): writes not at end of stream corrupt data
1745 [20292] dynamic-link: Comparison in elf/dl-open.c _dl_addr_inside_object
1747 [20311] nptl: please install proc_service.h
1748 [20366] libc: Compilation errors in installed headers in strict-compliance
1750 [20370] malloc: malloc: Arena free list management is still racy
1751 (incorrect fix in bug 19243)
1752 [20386] libc: assert (X = 0) does not result in GCC warning
1753 [20432] malloc: malloc: Minimize interface required for interposition
1754 [20435] libc: Missing unwind info in __startcontext causes infinite loop
1755 in _Unwind_Backtrace (CVE-2016-6323)
1756 [20444] hurd: recvmsg: PF_LOCAL sockets and msg_name lead to SIGLOST
1757 [20452] nptl: Addition of sysdep.o to libpthread.a breaks relinking
1759 [20455] math: [powerpc] fesetexceptflag fails to clear FE_INVALID
1760 [20459] localedata: et_EE: locale has wrong {p,n}_cs_precedes value
1761 [20477] network: resolv: incorrect double-checked locking related to
1763 [20478] libc: libc_ifunc macro and similar usages leads to false debug-
1765 [20495] math: x86_64 performance degradation due to AVX/SSE transition
1767 [20497] localedata: lt_LT: LC_TIME d_fmt used is obsolete
1768 [20508] dynamic-link: _dl_runtime_resolve_avx/_dl_runtime_profile_avx512
1769 cause transition penalty
1770 [20517] math: sparcv9 missing fdiml compat symbol
1771 [20524] manual: strverscmp is inconsistent
1772 [20525] libc: <sys/quota.h> should be based on kernel headers
1773 [20539] math: GCC 7 -static -lm fails to link at -O3
1774 [20554] libc: ld: bss-plt forced due to /usr/lib/libc_nonshared.a(ppc-
1776 [20558] string: POSIX bcopy/bzero decls do not implement Object Size
1778 [20591] network: Remove obsolete DNSSEC support
1779 [20592] network: DNS resource record type classification macros in
1780 <arpa/nameser.h> are incorrect
1781 [20593] network: Update DNS RR type definitions
1782 [20611] network: getaddrinfo accepts invalid numeric scope IDs
1783 [20615] build: glibc build fails when using --with-cpu=power9 --enable-
1785 [20629] network: libresolv: Remove support for bitstring labels
1787 [20647] libc: GLIBC quitting every program - glibc on Pentium-S leads to
1788 assertion: "maxidx >= 2"
1789 [20660] math: [arm] Use VSQRT
1790 [20662] libc: checking whether x86_64-pc-linux-gnu-gcc implicitly enables
1791 -fstack-protector no (32bit gcc 6.2.0 pie and ssp enable)
1792 [20689] libc: Test for FMA should also check for AVX.
1793 [20707] glob: gl_pathv entries not set to NULL with GLOB_DOOFFS
1794 [20715] math: iszero macro breaks existing code
1795 [20718] math: [powerpc] copysignl raises "invalid" for sNaN
1796 [20728] libc: powerpc: Missing TOC stub in clone
1797 [20729] build: glibc-2.24 fails to build for i486 with -Os
1798 [20750] build: Build fails with default PIE enabled gcc-6.x
1799 [20768] math: [alpha] sqrt fegetenv namespace
1800 [20785] libc: binutils 2.28 fails to assemble power6/memset.S file in
1802 [20787] math: float_t is defined as float incorrectly on x86_64 even if
1803 __FLT_EVAL_METHOD__ is 2
1804 [20790] network: rpcgen buffer overrun in get_prog_declaration
1805 [20822] nptl: race condition in __lll_unlock_elision on powerpc
1806 [20829] libc: crypt snprintf namespace
1807 [20847] libc: tst-vfork3 failure
1808 [20855] math: Default bits/mathdef.h has inappropriate float_t
1809 [20859] math: [sh4] FP_ILOGB0 invalid
1810 [20864] localedata: iconv: cp936 missing single-byte euro sign (0x80,
1811 U+20AC), not same as GBK
1812 [20915] dynamic-link: global-dynamic TLS broken on aarch64
1813 [20916] math: pow handling of sNaN arguments
1814 [20918] build: Building with --enable-nss-crypt fails tst-linkall-static.
1815 [20919] math: Bad pow (qNaN, 0) result with -lieee
1816 [20940] math: hypot sNaN handling
1817 [20947] math: fmax, fmin sNaN handling
1818 [20956] libc: debug/tst-backtrace3-6 don't work with -O3 anymore
1819 [20964] network: sunrpc: Stack-based buffer overflow in getrpcport with
1821 [20971] string: powerpc64/power7 memchr overflows internal pointer check
1822 [20973] nptl: robust mutexes: Lost wake-ups
1823 [20974] locale: bs_BA: yesexpr/noexpr regexes accept any character
1824 [20978] nis: strlen on null pointer in nss_nisplus
1825 [20985] nptl: robust mutexes: lowlevelrobustlock assembly on x86 blocks on
1827 [21014] string: i686 memchr overflows internal pointer check
1828 [21019] libc: [mips] n32 lseek incorrect on overflow
1829 [21022] libc: [microblaze] __backtrace get_frame_size namespace
1830 [21026] libc: [MIPS] readahead syscall is broken on n64
1831 [21028] math: Fallback fesetexceptflag should always succeed
1832 [21045] libc: [powerpc-nofpu] swapcontext does not restore signal mask
1833 [21047] math: arm: fpu_control.h: _FPU_GETCW/_FPU_SETCW is rejected by
1835 [21053] libc: [SH] Namespace pollution from sys/ucontext.h
1836 [21061] librt: [microblaze] librt lost clock_* exports
1837 [21073] libc: tunables: insecure environment variables passed to
1838 subprocesses with AT_SECURE
1839 [21081] string: Missing vzeroupper in memset-vec-unaligned-erms.S
1843 * The minimum Linux kernel version that this version of the GNU C Library
1844 can be used with is 3.2, except on i[4567]86 and x86_64, where Linux
1845 kernel version 2.6.32 or later suffices (on architectures that already
1846 required kernel versions more recent than 3.2, those requirements remain
1847 unchanged). Linux 3.2 or later kernel headers are required on all
1850 * The pap_AN locale has been deleted. This has been deprecated for a long
1851 time. It has been replaced by pap_AW & pap_CW, both of which have long
1852 been included in previous releases.
1854 * The readdir_r and readdir64_r functions have been deprecated. It is
1855 recommended to use readdir and readdir64 instead.
1857 * The type “union wait” has been removed. It was deprecated in the early
1858 1990s and never part of POSIX. Application code should use the int type
1859 instead of “union wait”.
1861 * A new NSS action is added to facilitate large distributed system
1862 administration. The action, MERGE, allows remote user stores like LDAP
1863 to be merged into local user stores like /etc/groups in order to provide
1864 easy to use, updated, and managed sets of merged credentials. The new
1865 action can be used by configuring it in /etc/nsswitch.conf:
1866 group: files [SUCCESS=merge] nis
1867 Implemented by Stephen Gallagher (Red Hat).
1869 * The deprecated __malloc_initialize_hook variable has been removed from the
1872 * The long unused localedef --old-style option has been removed. It hasn't
1873 done anything in over 16 years. Scripts using this option can safely
1876 * nextupl, nextup, nextupf, nextdownl, nextdown and nextdownf are added to
1877 libm. They are defined by TS 18661 and IEEE754-2008. The nextup functions
1878 return the next representable value in the direction of positive infinity
1879 and the nextdown functions return the next representable value in the
1880 direction of negative infinity. These are currently enabled as GNU
1883 Security related changes:
1885 * An unnecessary stack copy in _nss_dns_getnetbyname_r was removed. It
1886 could result in a stack overflow when getnetbyname was called with an
1887 overly long name. (CVE-2016-3075)
1889 * Previously, getaddrinfo copied large amounts of address data to the stack,
1890 even after the fix for CVE-2013-4458 has been applied, potentially
1891 resulting in a stack overflow. getaddrinfo now uses a heap allocation
1892 instead. Reported by Michael Petlan. (CVE-2016-3706)
1894 * The glob function suffered from a stack-based buffer overflow when it was
1895 called with the GLOB_ALTDIRFUNC flag and encountered a long file name.
1896 Reported by Alexander Cherepanov. (CVE-2016-1234)
1898 * The Sun RPC UDP client could exhaust all available stack space when
1899 flooded with crafted ICMP and UDP messages. Reported by Aldy Hernandez'
1900 alloca plugin for GCC. (CVE-2016-4429)
1902 * The IPv6 name server management code in libresolv could result in a memory
1903 leak for each thread which is created, performs a failing naming lookup,
1904 and exits. Over time, this could result in a denial of service due to
1905 memory exhaustion. Reported by Matthias Schiffer. (CVE-2016-5417)
1907 The following bugs are resolved with this release:
1909 [1170] localedata: ne_NP: update Nepali locale definition file
1910 [3629] manual: stpcpy description in string.texi refers to MS-DOG instead
1912 [6527] malloc: [powerpc] Malloc alignment insufficient for PowerPC
1913 [6796] math: fdim() does not set errno on overflow
1914 [10354] libc: posix_spawn should use vfork() in more cases than presently
1915 [11213] localedata: localedata: add copyright disclaimer to locale files
1916 [12143] localedata: chr_US: new Cherokee locale
1917 [12450] localedata: sgs_LT: new locale
1918 [12676] localedata: ln_CD: new locale
1919 [13237] localedata: LC_ADDRESS.country_name: update all locales w/latest
1921 [13304] math: fma, fmaf, fmal produce wrong results
1922 [14259] build: --localedir arg to configure is ignored
1923 [14499] nptl: Does posix_spawn invoke atfork handlers / use vfork?
1924 [14750] libc: Race condition in posix_spawn vfork usage vs signal handlers
1925 [14934] localedata: es_CL: wrong first weekday chilean locale
1926 [15262] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of
1928 [15263] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of 1/0
1930 [15264] localedata: LC_MESSAGES.yesstr/nostr: lacking in many locales
1931 [15368] nptl: raise() is not async-signal-safe
1932 [15479] math: ceil, floor, round and trunc raise inexact exception
1933 [15578] localedata: kk_KZ: various updates
1934 [16003] localedata: pap_AN: punt old locale
1935 [16137] localedata: iw_IL: punt old locale
1936 [16190] localedata: eo: new esperanto locale
1937 [16374] localedata: lv_LV: change currency symbol in LC_MONETARY to euro
1938 [16742] malloc: race condition: pthread_atfork() called before first
1939 malloc() results in unexpected locking behaviour/deadlocks
1940 [16975] localedata: LC_MESSAGES.yesexpr/noexpr: revisit capitalization in
1942 [16983] localedata: postal_fmt does not allow %l and %n modifiers
1943 [17565] localedata: pt_PT: wrong (work-)week start
1944 [17899] math: [powerpc] floorl returns negative zero with FE_DOWNWARD
1945 [17950] build: Build fails with -msse
1946 [18205] localedata: be_BY*: wrong first_weekday and first_workday
1947 [18433] libc: posix_spawn does not return correctly upon failure to
1949 [18453] localedata: charmaps/IBM875: incorrect codes
1950 [18712] string: bits/string2.h incompatible with -O2 -Werror=packed
1952 [18896] localedata: he_IL: improvements for currency
1953 [18911] localedata: ro_RO: Correcting week day name for "Tuesday" in
1954 Romanian locale data
1955 [18960] locale: s390: _nl_locale_subfreeres uses larl opcode on misaligned
1957 [19056] libc: Deprecate readdir_r
1958 [19133] localedata: pt_*: days & months should be lowercase in Portuguese
1960 [19198] localedata: nl_NL: small improvements for Dutch locales
1961 [19257] network: Per-thread memory leak in __res_vinit with IPv6
1962 nameservers (CVE-2016-5417)
1963 [19269] build: tst-audit4 and tst-audit10 failures with gcc-6 on non avx
1965 [19400] locale: Language missing in "iso-639.def", trivial fix in
1967 [19431] malloc: Deadlock between fflush, getdelim, and fork
1968 [19505] libc: Incorrect file descriptor validity checks in
1969 posix_spawn_file_actions_add{open,close,dup2}
1970 [19509] dynamic-link: dlsym, dlvsym do not report errors through dlerror
1971 when using RTLD_NEXT
1972 [19512] locale: Stale `#ifndef HAVE_BUILTIN_EXPECT' in
1973 `intl/{gettextP,loadinfo}.h'
1974 [19534] libc: execle, execlp may use malloc
1975 [19568] localedata: *_CH: Swiss locales have inconsistent start of week
1976 [19573] network: res_nclose and __res_maybe_init disagree about name
1977 server initialization, breaking Hesiod
1978 [19575] localedata: Status of GB18030 tables
1979 [19581] localedata: sr_* date_fmt string contains additional newline
1980 [19583] string: SSSE3_Fast_Copy_Backward flag needs to be enabled for AMD
1982 [19592] math: [ldbl-128ibm] ceill incorrect in non-default rounding modes
1983 [19593] math: [ldbl-128ibm] truncl incorrect in non-default rounding modes
1984 [19594] math: [ldbl-128ibm] roundl incorrect in non-default rounding modes
1985 [19595] math: [ldbl-128ibm] fmodl incorrect for results in subnormal
1987 [19602] math: [ldbl-128ibm] fmodl handling of equal arguments with low
1989 [19603] math: [ldbl-128ibm] remainderl, remquol incorrect sign handling in
1991 [19610] dynamic-link: ldconfig -X removes stale symbolic links
1992 [19613] libc: s390x (64 bit) macro expansion WCOREDUMP and others
1993 [19633] locale: strfmon_l applies global locale to number formatting
1994 [19642] network: Memory leak in getnameinfo
1995 [19648] libc: test-skeleton.c: Do not set RLIMIT_DATA
1996 [19653] libc: Potential for NULL pointer dereference (CWE-476) in
1998 [19654] math: [x86_64] Need testcase for BZ #19590 fix
1999 [19671] localedata: Missing Sanity Check for malloc() in 'tst-fmon.c' &
2001 [19674] math: [ldbl-128ibm] powl incorrect overflow handling
2002 [19677] math: [ldbl-128ibm] remainderl equality test incorrect for zero
2004 [19678] math: [ldbl-128ibm] nextafterl, nexttowardl incorrect sign of zero
2006 [19679] dynamic-link: gcc-4.9.3 C++ exception handling broken due to
2008 [19726] locale: Converting UCS4LE to INTERNAL with iconv() does not update
2009 pointers and lengths in error-case.
2010 [19727] locale: Converting from/to UTF-xx with iconv() does not always
2011 report errors on UTF-16 surrogates values.
2012 [19755] nscd: nscd assertion failure in gc
2013 [19758] dynamic-link: Typo in EXTRA_LD_ENVVARS for x86-64
2014 [19759] libc: mempcpy shouldn't be inlined
2015 [19762] dynamic-link: HAS_CPU_FEATURE/HAS_ARCH_FEATURE are easy to misuse
2016 [19765] libc: s390 needs an optimized mempcpy
2017 [19779] glob: glob: buffer overflow with GLOB_ALTDIRFUNC due to incorrect
2018 NAME_MAX limit assumption (CVE-2016-1234)
2019 [19783] build: benchtests don't support --enable-hardcoded-path-in-tests
2020 [19787] network: Missing and incorrect truncation checks in getnameinfo
2021 [19790] math: [ldbl-128ibm] nearbyintl incorrect in non-default rounding
2023 [19791] network: Assertion failure in res_query.c with un-connectable name
2025 [19792] libc: MIPS: backtrace yields infinite backtrace with makecontext
2026 [19822] math: libm.so install clobbers old version
2027 [19825] network: resolv: send_vc can return uninitialized data in second
2028 response to getaddrinfo
2029 [19830] network: nss_dns: should check RDATA length against buffer length
2030 [19831] network: nss_dns: getaddrinfo returns uninitialized data when
2031 confronted with A/AAAA records of invalid size
2032 [19837] nss: nss_db: No retries for some long lines with a larger buffer
2033 [19848] math: powl(10,n) for n=-4,-5,-6,-7 is off by more than 1 ULP
2034 [19853] stdio: Printing IBM long double in decimal with high precision is
2036 [19860] build: x86_64: compile errors for tst-audit10 and tst-auditmod10b
2037 [19861] nptl: libpthread IFUNC resolver for fork can lead to crash
2038 [19862] network: resolv, nss_dns: Remove remaining logging of unexpected
2040 [19865] network: Assertion failure or memory leak in
2041 _nss_dns_getcanonname_r
2042 [19868] network: nss_dns: netent code does not skip over non-PTR records
2043 [19879] network: nss_dns: Stack overflow in getnetbyname implementation
2045 [19881] string: Improve x86-64 memset
2046 [19907] string: Incorrect memcpy tests
2047 [19916] dynamic-link: S390: fprs/vrs are not saved/restored while
2049 [19925] libc: termios.h XCASE namespace
2050 [19928] string: memmove-vec-unaligned-erms.S is slow with large data size
2051 [19929] libc: limits.h NL_NMAX namespace
2052 [19931] stdio: Memory leak in vfprintf
2053 [19957] libc: clone(CLONE_VM) access invalid parent memory
2054 [19963] localedata: en_IL: New locale
2055 [19989] stdio: stdio.h cuserid namespace
2056 [19994] network: getaddrinfo does not restore RES_USE_INET6 flag in
2058 [19996] locale: langinfo.h nl_langinfo_l namespace
2059 [20005] stdio: fflush on a file opened with fmemopen resets position to 0
2060 [20010] network: getaddrinfo: Stack overflow in hostent translation
2062 [20012] stdio: libio: fmemopen append mode failure
2063 [20014] stdio: stdio.h namespace for pre-threads POSIX
2064 [20017] network: resolv: Use gmtime_r instead of gmtime in p_secstodate
2065 [20023] libc: fcntl.h timespec namespace
2066 [20024] math: [x86_64] vectorized sincos trashes the stack
2067 [20031] network: nss_hesiod: Heap overflow in get_txt_records
2068 [20041] time: sys/time.h timespec namespace
2069 [20043] libc: unistd.h missing cuserid for UNIX98 and before
2070 [20044] libc: unistd.h missing pthread_atfork for UNIX98
2071 [20051] libc: ttyslot in wrong header under wrong conditions
2072 [20054] libc: gethostname not declared for XPG4
2073 [20055] libc: termios.h missing tcgetsid for XPG4
2074 [20072] dynamic-link: x86 init_cpu_features is called twice in static
2076 [20073] libc: sys/stat.h fchmod namespace
2077 [20074] libc: stdlib.h rand_r namespace
2078 [20076] libc: sys/stat.h missing S_IFSOCK, S_ISSOCK for XPG4
2079 [20094] libc: stdlib.h should not declare grantpt, ptsname, unlockpt for
2081 [20111] libc: struct sockaddr_storage cannot be aggregate-copied
2082 [20112] network: sunrpc: stack (frame) overflow in Sun RPC clntudp_call
2084 [20115] string: Extra alignment in memset-vec-unaligned-erms.S
2085 [20119] libc: Wrong mask for processors level type from CPUID
2086 [20139] dynamic-link: Upper part of zmm is zeroed if Glibc is built with
2087 AS not supporting AVX512
2088 [20151] math: [ldbl-128/ldbl-128ibm] j0l, j1l, y0l, y1l return sNaN for
2090 [20153] math: [ldbl-128ibm] sqrtl (sNaN) returns sNaN
2091 [20156] math: [ldbl-128ibm] ceill, rintl etc. return sNaN for sNaN
2093 [20157] math: [powerpc] fabsl (sNaN) wrongly raises "invalid"
2094 [20160] math: [powerpc] ceil, rint etc. return sNaN for sNaN input
2095 [20178] libc: posix_spawn{p} should not call exit
2096 [20191] stdio: libio: vtables hardening
2097 [20195] string: FMA4 detection requires CPUID execution with register
2099 [20198] libc: quick_exit incorrectly destroys C++11 thread objects.
2100 [20205] math: [i386/x86_64] nextafterl incorrect incrementing negative
2102 [20212] math: acos (sNaN) returns sNaN
2103 [20213] math: asin (sNaN) returns sNaN
2104 [20214] network: Linux header sync with linux/in6.h and ipv6.h again.
2105 [20218] math: [i386] asinhl (sNaN) returns sNaN
2106 [20219] math: [i386] atanhl (sNaN) returns sNaN
2107 [20222] stdio: fopencookie: Mangle function pointers
2108 [20224] math: [i386] cbrtl (sNaN) returns sNaN
2109 [20225] math: ldexp, scalbn, scalbln return sNaN for sNaN input
2110 [20226] math: [i386/x86_64] expl, exp10l, expm1l return sNaN for sNaN
2112 [20227] math: [i386/x86_64] logl (sNaN) returns sNaN
2113 [20228] math: [i386/x86_64] log10l (sNaN) returns sNaN
2114 [20229] math: [i386/x86_64] log1pl (sNaN) returns sNaN
2115 [20232] math: [ldbl-128] expm1l (sNaN) returns sNaN
2116 [20233] math: [ldbl-128ibm] expm1l (sNaN) returns sNaN
2117 [20234] math: [ldbl-128ibm] log1pl (sNaN) returns sNaN
2118 [20235] math: [i386/x86_64] log2l (sNaN) returns sNaN
2119 [20237] nss: nss_db: get*ent segfaults without preceding set*ent
2120 [20240] math: modf (sNaN) returns sNaN
2121 [20248] libc: debug/tst-longjump_chk2 calls printf from a signal handler
2122 [20250] math: frexp (sNaN) returns sNaN
2123 [20252] math: atan2 (sNaN, qNaN) fails to raise "invalid"
2124 [20255] math: [i386] fdim, fdimf return with excess range and precision /
2126 [20256] math: [i386/x86_64] fdiml returns sNaN for sNaN input
2127 [20260] string: ../sysdeps/x86/bits/string.h:1092:3: error: array
2128 subscript is below array bounds [-Werror=array-bounds]
2129 [20262] nis: _nss_nis_initgroups_dyn always returns NSS_STATUS_NOTFOUND
2130 [20263] nptl: robust mutex deadlocks if other thread requests timedlock
2132 [20277] libc: $dp is not initialized correctly in sysdeps/hppa/start.S
2133 [20284] malloc: malloc: Corrupt arena avoidance causes unnecessary mmap
2135 [20296] math: [i386/x86_64] scalbl returns sNaN for sNaN input, missing
2136 "invalid" exceptions
2137 [20314] nptl: make[4]: *** [/usr/include/stdlib.h] Error 1
2138 [20316] localedata: id_ID: Februari instead of Pebruari
2139 [20327] string: POWER8 strcasecmp returns incorrect result
2140 [20347] math: Failure: Test: j0_downward (0xap+0)
2141 [20348] libc: FAIL: misc/tst-preadvwritev64
2142 [20349] libc: 64-bit value is passed differently in p{readv,writev}{64}
2143 [20350] libc: There is no test for p{read,write}64
2144 [20357] math: Incorrect cos result for 1.5174239687223976
2145 [20384] build: Don't run libmvec-sincos-avx* tests on non avx machines
2149 * Unicode 8.0.0 Support: Character encoding, character type info, and
2150 transliteration tables are all updated to Unicode 8.0.0, using new
2151 and/or improved generator scripts contributed by Mike FABIAN (Red Hat).
2152 These updates cause user visible changes, such as the fixes for bugs
2153 89, 16061, and 18568.
2155 * sched_setaffinity, pthread_setaffinity_np no longer attempt to guess the
2156 kernel-internal CPU set size. This means that requests that change the
2157 CPU affinity which failed before (for example, an all-ones CPU mask) will
2158 now succeed. Applications that need to determine the effective CPU
2159 affinities need to call sched_getaffinity or pthread_getaffinity_np after
2160 setting it because the kernel can adjust it (and the previous size check
2161 would not detect this in the majority of cases).
2163 * The fts.h header can now be used with -D_FILE_OFFSET_BITS=64. With LFS
2164 the following new symbols are used: fts64_children, fts64_close,
2165 fts64_open, fts64_read and fts64_set.
2167 * getaddrinfo now detects certain invalid responses on an internal netlink
2168 socket. If such responses are received, an affected process will
2169 terminate with an error message of "Unexpected error <number> on netlink
2170 descriptor <number>" or "Unexpected netlink response of size <number> on
2171 descriptor <number>". The most likely cause for these errors is a
2172 multi-threaded application which erroneously closes and reuses the netlink
2173 file descriptor while it is used by getaddrinfo.
2175 * A defect in the malloc implementation, present since glibc 2.15 (2012) or
2176 glibc 2.10 via --enable-experimental-malloc (2009), could result in the
2177 unnecessary serialization of memory allocation requests across threads.
2178 The defect is now corrected. Users should see a substantial increase in
2179 the concurent throughput of allocation requests for applications which
2180 trigger this bug. Affected applications typically create and
2181 destroy threads frequently. (Bug 19048 was reported and analyzed by
2184 * There is now a --disable-timezone-tools configure option for disabling the
2185 building and installing of the timezone related utilities (zic, zdump, and
2186 tzselect). This is useful for people who build the timezone data and code
2187 independent of the GNU C Library.
2189 * The obsolete header <regexp.h> has been removed. Programs that require
2190 this header must be updated to use <regex.h> instead.
2192 * The obsolete functions bdflush, create_module, get_kernel_syms,
2193 query_module and uselib are no longer available to newly linked binaries;
2194 the header <sys/kdaemon.h> has been removed. These functions and header
2195 were specific to systems using the Linux kernel and could not usefully be
2196 used with the GNU C Library on systems with version 2.6 or later of the
2199 * Optimized string, wcsmbs and memory functions for IBM z13.
2200 Implemented by Stefan Liebler.
2202 * Newly linked programs that define a variable called signgam will no longer
2203 have it set by the lgamma, lgammaf and lgammal functions. Programs that
2204 require signgam to be set by those functions must ensure that they use the
2205 variable provided by the GNU C Library and declared in <math.h>, without
2206 defining their own copy.
2208 * The minimum GCC version that can be used to build this version of the GNU
2209 C Library is GCC 4.7. Older GCC versions, and non-GNU compilers, can
2210 still be used to compile programs using the GNU C Library.
2212 Security related changes:
2214 * An out-of-bounds value in a broken-out struct tm argument to strftime no
2215 longer causes a crash. Reported by Adam Nielsen. (CVE-2015-8776)
2217 * The LD_POINTER_GUARD environment variable can no longer be used to disable
2218 the pointer guard feature. It is always enabled. Previously,
2219 LD_POINTER_GUARD could be used to disable security hardening in binaries
2220 running in privileged AT_SECURE mode. Reported by Hector Marco-Gisbert.
2223 * An integer overflow in hcreate and hcreate_r could lead to an
2224 out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
2226 * The catopen function no longer has unbounded stack usage. Reported by
2227 Max. (CVE-2015-8779)
2229 * The nan, nanf and nanl functions no longer have unbounded stack usage
2230 depending on the length of the string passed as an argument to the
2231 functions. Reported by Joseph Myers. (CVE-2014-9761)
2233 * A stack-based buffer overflow was found in libresolv when invoked from
2234 libnss_dns, allowing specially crafted DNS responses to seize control
2235 of execution flow in the DNS client. The buffer overflow occurs in
2236 the functions send_dg (send datagram) and send_vc (send TCP) for the
2237 NSS module libnss_dns.so.2 when calling getaddrinfo with AF_UNSPEC
2238 family. The use of AF_UNSPEC triggers the low-level resolver code to
2239 send out two parallel queries for A and AAAA. A mismanagement of the
2240 buffers used for those queries could result in the response of a query
2241 writing beyond the alloca allocated buffer created by
2242 _nss_dns_gethostbyname4_r. Buffer management is simplified to remove
2243 the overflow. Thanks to the Google Security Team and Red Hat for
2244 reporting the security impact of this issue, and Robert Holiday of
2245 Ciena for reporting the related bug 18665. (CVE-2015-7547)
2247 The following bugs are resolved with this release:
2249 [89] localedata: Locales nb_NO and nn_NO should transliterate æøå
2250 [887] math: Math library function "logb" and "nextafter" inconsistent
2251 [2542] math: Incorrect return from float gamma (-0X1.FA471547C2FE5P+1)
2252 [2543] math: Incorrect return from float gamma (-0X1.9260DCP+1)
2253 [2558] math: Incorrect return from double gamma (-0X1.FA471547C2FE5P+1)
2254 [2898] libc: [improve] warning: the use of `mktemp' is dangerous, better
2256 [4404] localedata: German translation of "Alarm clock" is misleading
2257 [6799] math: nextafter() and nexttoward() doen't set errno on
2258 overflow/underflow errors
2259 [6803] math: scalb(), scalbln(), scalbn() do not set errno on
2261 [10432] nis: _nss_nis_setnetgrent assertion failure
2262 [11460] libc: fts has no LFS support
2263 [12926] network: getaddrinfo()/make_request() may spin forever
2264 [13065] nptl: Race condition in pthread barriers
2265 [13690] nptl: pthread_mutex_unlock potentially cause invalid access
2266 [14341] dynamic-link: Dynamic linker crash when DT_JMPREL and DT_REL{,A}
2268 [14551] math: [ldbl-128ibm] strtold overflow handling for IBM long double
2269 [14912] libc: Rename non-installed bits/*.h headers
2270 [15002] libc: Avoid undefined behavior in posix_fallocate overflow check
2271 [15367] math: Let gcc use __builtin_isinf
2272 [15384] math: One constant fewer in ieee754/dbl-64/wordsize-64/s_finite.c
2273 [15421] math: lgamma wrongly sets signgam for ISO C
2274 [15470] math: [arm] On ARM llrintl() and llroundl() do not raise
2275 FE_INVALID with argument out of range
2276 [15491] math: [i386/x86_64] x86 nearbyint implementations wrongly clear
2278 [15786] dynamic-link: ifunc resolver functions can smash function
2280 [15918] math: Unnecessary check for equality in hypotf()
2281 [16061] localedata: Review / update transliteration data
2282 [16068] math: [i386/x86_64] x86 and x86_64 fesetenv exclude state they
2284 [16141] time: strptime %z offset restriction
2285 [16171] math: drem should be alias of remainder
2286 [16296] math: fegetround is pure?
2287 [16347] math: [ldbl-128ibm] ldbl-128/e_lgammal_r.c may not be suitable.
2288 [16364] libc: sleep may leave SIGCHLD blocked on sync cancellation on
2290 [16399] math: [mips] lrint / llrint / lround / llround missing exceptions
2291 [16415] math: Clean up ldbl-128 / ldbl-128ibm expm1l for large positive
2293 [16422] math: [powerpc] math-float, math-double failing llrint tests with
2294 "Exception "Inexact" set" on ppc32
2295 [16495] localedata: nl_NL: date_fmt: shuffle year/month around
2296 [16517] math: Missing underflow exception from tanf/tan/tanl
2297 [16519] math: Missing underflow exception from sinhf
2298 [16520] math: Missing underflow exception from tanhf
2299 [16521] math: Missing underflow exception from exp2
2300 [16620] math: [ldbl-128ibm] exp10l spurious overflows / bad directed
2302 [16734] stdio: fopen calls mmap to allocate its buffer
2303 [16961] math: nan function incorrect handling of bad sequences
2304 [16962] math: nan function unbounded stack allocation (CVE-2014-9761)
2305 [16973] localedata: Fix lang_lib/lang_term as per ISO 639-2
2306 [16985] locale: localedef: confusing error message when opening output
2308 [17118] math: ctanh(INFINITY + 2 * I) returns incorrect value
2309 [17197] locale: Redundant shift character in iconv conversion output at
2311 [17243] libc: trunk/posix/execl.c:53: va_args problem ?
2312 [17244] libc: trunk/sysdeps/unix/sysv/linux/semctl.c:116: va_args muxup ?
2313 [17250] dynamic-link: static linking breaks nss loading
2314 (getaddrinfo/getpwnam/etc...)
2315 [17404] libc: atomic_exchange_rel lacking a barrier on MIPS16, GCC before
2317 [17441] math: isnan() should use __builtin_isnan() in GCC
2318 [17514] nptl: Assert failure unlocking ERRORCHECK mutex after timedlock
2319 (related to lock elision)
2320 [17787] manual: Exponent on page 324 of the PDF ends prematurely
2321 [17886] time: strptime should be able to parse "Z" as a timezone with %z
2322 [17887] time: strptime should be able to parse "+01:00" style timezones
2323 [17905] libc: catopen() Multiple unbounded stack allocations
2325 [18084] libc: backtrace (..., 0) dumps core on x86
2326 [18086] libc: nice() sets errno to 0 on success
2327 [18240] libc: hcreate, hcreate_r should fail with ENOMEM if element count
2328 is too large (CVE-2015-8778)
2329 [18251] dynamic-link: SONAME missing when audit modules provides path
2330 [18265] libc: add attributes for wchar string and memory functions
2331 [18370] math: csqrt missing underflows
2332 [18421] libc: [hppa] read-only segment has dynamic relocations
2333 [18472] libc: Obsolete syscall wrappers should be compat symbols
2334 [18480] libc: hppa glibc miscompilation in sched_setaffinity()
2335 [18491] localedata: Update tr_TR LC_CTYPE as part of Unicode updates
2336 [18525] localedata: Remove locale timezone information
2337 [18560] libc: [powerpc] spurious bits/ipc.h definitions
2338 [18568] localedata: Update locale data to Unicode 8.0
2339 [18589] locale: sort-test.sh fails at random
2340 [18595] math: ctan, ctanh missing underflows
2341 [18604] libc: assert macro-expands its argument
2342 [18610] math: S390: fetestexcept() reports any exception if DXC-code
2343 contains a vector instruction exception.
2344 [18611] math: j1, jn missing errno setting on underflow
2345 [18618] localedata: sync Chechen locale definitions with other *_RU
2347 [18647] math: powf(-0x1.000002p0, 0x1p30) returns 0 instead of +inf
2348 [18661] libc: Some x86-64 assembly codes don't align stack to 16 bytes
2349 [18665] network: In send_dg, the recvfrom function is NOT always using the
2350 buffer size of a newly created buffer (CVE-2015-7547)
2351 [18674] libc: [i386] trunk/sysdeps/i386/tst-auditmod3b.c:84: possible
2353 [18675] libc: fpathconf(_PC_NAME_MAX) fails against large filesystems for
2355 [18681] libc: regexp.h is obsolete and buggy, and should be desupported
2356 [18699] math: tilegx cproj() for various complex infinities does not yield
2358 [18724] libc: Harden put*ent functions against data injection
2359 [18743] nptl: PowerPC: findutils testcase fails with --enable-lock-elision
2360 [18755] build: build errors with -DNDEBUG
2361 [18757] stdio: fmemopen fails to set errno on failure
2362 [18778] dynamic-link: ld.so crashes if failed dlopen causes libpthread to
2364 [18781] libc: openat64 lacks O_LARGEFILE
2365 [18787] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/atomic.h:71:6:
2366 error: can’t find a register in class ‘R1_REGS’ while reloading ‘asm’
2367 [18789] math: [ldbl-128ibm] sinhl inaccurate near 0
2368 [18790] math: [ldbl-128ibm] tanhl inaccurate
2369 [18795] libc: stpncpy fortification misses buffer lengths that are
2370 statically too large
2371 [18796] build: build fails for --disable-mathvec
2372 [18803] math: hypot missing underflows
2373 [18820] stdio: fmemopen may leak memory on failure
2374 [18823] math: csqrt spurious underflows
2375 [18824] math: fma spurious underflows
2376 [18825] math: pow missing underflows
2377 [18857] math: [ldbl-128ibm] nearbyintl wrongly uses signaling comparisons
2378 [18868] nptl: pthread_barrier_init typo has in-theory-undefined behavior
2379 [18870] build: sem_open.c fails to compile with missing symbol
2381 [18872] stdio: Fix memory leak in printf_positional
2382 [18873] libc: posix_fallocate overflow check ineffective
2383 [18875] math: Excess precision leads incorrect libm
2384 [18877] libc: arm: mmap offset regression
2385 [18887] libc: memory corruption when using getmntent on blank lines
2386 [18918] localedata: hu_HU: change time to HH:MM:SS format
2387 [18921] libc: Regression: extraneous stat() and fstat() performed by
2389 [18928] dynamic-link: LD_POINTER_GUARD is not ignored for privileged
2390 binaries (CVE-2015-8777)
2391 [18951] math: tgamma missing underflows
2392 [18952] math: [ldbl-128/ldbl-128ibm] lgammal spurious "invalid", incorrect
2394 [18953] localedata: lt_LT: change currency symbol to the euro
2395 [18956] math: powf inaccuracy
2396 [18961] math: [i386] exp missing underflows
2397 [18966] math: [i386] exp10 missing underflows
2398 [18967] math: math.h XSI POSIX namespace (gamma, isnan, scalb)
2399 [18969] build: multiple string test failures due to missing locale
2401 [18970] libc: Reference of pthread_setcancelstate in libc.a
2402 [18977] math: float / long double Bessel functions not in XSI POSIX
2403 [18980] math: i386 libm functions return with excess range and precision
2404 [18981] math: i386 scalb*, ldexp return with excess range and precision
2405 [18982] stdio: va_list and vprintf
2406 [18985] time: Passing out of range data to strftime() causes a segfault
2408 [19003] math: [x86_64] fma4 version of pow inappropriate contraction
2409 [19007] libc: FAIL: elf/check-localplt with -z now and binutils 2.26
2410 [19012] locale: iconv_open leaks memory on error path
2411 [19016] math: clog, clog10 inaccuracy
2412 [19018] nptl: Mangle function pointers in tls_dtor_list
2413 [19032] math: [i386] acosh (-qNaN) spurious "invalid" exception
2414 [19046] math: ldbl-128 / ldbl-128ibm lgamma bad overflow handling
2415 [19048] malloc: malloc: arena free list can become cyclic, increasing
2417 [19049] math: [powerpc] erfc incorrect zero sign
2418 [19050] math: [powerpc] log* incorrect zero sign
2419 [19058] math: [x86_64] Link fail with -fopenmp and -flto
2420 [19059] math: nexttoward overflow incorrect in non-default rounding modes
2421 [19071] math: ldbl-96 lroundl incorrect just below powers of 2
2422 [19074] network: Data race in _res_hconf_reorder_addrs
2423 [19076] math: [ldbl-128ibm] log1pl (-1) wrong sign of infinity
2424 [19077] math: [ldbl-128ibm] logl (1) incorrect sign of zero result
2425 [19078] math: [ldbl-128ibm] expl overflow incorrect in non-default
2427 [19079] math: dbl-64/wordsize-64 lround based on llround incorrect for
2429 [19085] math: ldbl-128 lrintl, lroundl missing exceptions for 32-bit long
2430 [19086] manual: posix_fallocate64 documented argument order is wrong.
2431 [19088] math: lround, llround missing exceptions close to overflow
2433 [19094] math: lrint, llrint missing exceptions close to overflow threshold
2434 [19095] math: dbl-64 lrint incorrect for 64-bit long
2435 [19122] dynamic-link: Unnecessary PLT relocations in librtld.os
2436 [19124] dynamic-link: ld.so failed to build with older assmebler
2437 [19125] math: [powerpc32] llroundf, llround incorrect exceptions
2438 [19129] dynamic-link: [arm] Concurrent lazy TLSDESC resolution can crash
2439 [19134] math: [powerpc32] lround, lroundf spurious exceptions
2440 [19137] libc: i386/epoll_pwait.S doesn't support cancellation
2441 [19143] nptl: Remove CPU set size checking from sched_setaffinity,
2442 pthread_setaffinity_np
2443 [19156] math: [ldbl-128] j0l spurious underflows
2444 [19164] nptl: tst-getcpu fails with many possible CPUs
2445 [19168] math: math/test-ildoubl and math/test-ldouble failure
2446 [19174] nptl: PowerPC: TLE enabled pthread mutex performs poorly.
2447 [19178] dynamic-link: ELF_RTYPE_CLASS_EXTERN_PROTECTED_DATA confuses
2449 [19181] math: [i386/x86_64] fesetenv (FE_DFL_ENV), fesetenv
2450 (FE_NOMASK_ENV) do not clear SSE exceptions
2451 [19182] malloc: malloc deadlock between ptmalloc_lock_all and
2452 _int_new_arena/reused_arena
2453 [19189] math: [ldbl-128] log1pl (-qNaN) spurious "invalid" exception
2454 [19201] math: dbl-64 remainder incorrect sign of zero result
2455 [19205] math: bits/math-finite.h conditions do not match math.h and
2457 [19209] math: bits/math-finite.h wrongly maps ldexp to scalbn
2458 [19211] math: lgamma functions do not set signgam for -ffinite-math-only
2459 for C99-based standards
2460 [19212] libc: features.h not -Wundef clean
2461 [19213] math: [i386/x86_64] log* (1) incorrect zero sign for -ffinite-
2463 [19214] libc: Family and model identification for AMD CPU's are incorrect.
2464 [19219] libc: GLIBC build fails for ia64 with missing __nearbyintl
2465 [19228] math: [powerpc] nearbyint wrongly clears "inexact", leaves traps
2467 [19235] math: [powerpc64] lround, lroundf, llround, llroundf spurious
2468 "inexact" exceptions
2469 [19238] math: [powerpc] round, roundf spurious "inexact" for integer
2471 [19242] libc: strtol incorrect in Turkish locales
2472 [19243] malloc: reused_arena can pick an arena on the free list, leading
2473 to an assertion failure and reference count corruption
2474 [19253] time: tzset() ineffective when temporary TZ did not include DST
2476 [19266] math: strtod ("NAN(I)") incorrect in Turkish locales
2477 [19270] math: [hppa] Shared libm missing __isnanl
2478 [19285] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/mman.h: missing
2479 MAP_HUGETLB and MAP_STACK defines
2480 [19313] nptl: Wrong __cpu_mask for x32
2481 [19347] libc: grantpt: try to force a specific gid even without pt_chown
2482 [19349] math: [ldbl-128ibm] tanhl inaccurate for small arguments
2483 [19350] math: [ldbl-128ibm] sinhl spurious overflows
2484 [19351] math: [ldbl-128ibm] logl inaccurate near 1
2485 [19363] time: x32: times() return value wrongly truncates/sign extends
2487 [19367] dynamic-link: Improve branch prediction on Silvermont
2488 [19369] network: Default domain name not reset by res_ninit when "search"
2489 / "domain" entry is removed from resolv.conf
2490 [19375] math: powerpc: incorrect results for POWER7 logb with negative
2492 [19385] localedata: bg_BG: time separator should be colon, not comma
2493 [19408] libc: linux personality syscall wrapper may erroneously return an
2494 error on 32-bit architectures
2495 [19415] libc: dladdr returns wrong names on hppa
2496 [19432] libc: iconv rejects redundant escape sequences in IBM900, IBM903,
2497 IBM905, IBM907, and IBM909
2498 [19439] math: Unix98 isinf and isnan functions conflict with C++11
2499 [19443] build: build failures with -DDEBUG
2500 [19451] build: Make check fails on test-double-vlen2
2501 [19462] libc: Glibc failed to build with -Os
2502 [19465] math: Wrong code with -Os
2503 [19466] time: time/tst-mktime2.c is compiled into an infinite loop with
2505 [19467] string: Fast_Unaligned_Load needs to be enabled for Excavator core
2507 [19475] libc: Glibc 2.22 doesn't build on sparc [PATCH]
2508 [19486] math: S390: Math tests fail with "Exception Inexact set".
2509 [19529] libc: [ARM]: FAIL: stdlib/tst-makecontext
2510 [19550] libc: [mips] mmap negative offset handling inconsistent with other
2512 [19590] math: Fail to build shared objects that use libmvec.so functions.
2516 * The following bugs are resolved with this release:
2518 438, 2981, 4719, 6544, 6792, 11216, 12836, 13028, 13064, 13151, 13152,
2519 14094, 14113, 14292, 14841, 14906, 14958, 15319, 15467, 15790, 15969,
2520 16159, 16339, 16350, 16351, 16352, 16353, 16361, 16512, 16526, 16538,
2521 16559, 16560, 16704, 16783, 16850, 17053, 17090, 17195, 17269, 17293,
2522 17322, 17403, 17475, 17523, 17542, 17569, 17581, 17588, 17596, 17620,
2523 17621, 17628, 17631, 17692, 17711, 17715, 17776, 17779, 17792, 17833,
2524 17836, 17841, 17912, 17916, 17930, 17932, 17944, 17949, 17964, 17965,
2525 17967, 17969, 17977, 17978, 17987, 17991, 17996, 17998, 17999, 18007,
2526 18019, 18020, 18029, 18030, 18032, 18034, 18036, 18038, 18039, 18042,
2527 18043, 18046, 18047, 18049, 18068, 18078, 18080, 18093, 18100, 18104,
2528 18110, 18111, 18116, 18125, 18128, 18134, 18138, 18185, 18196, 18197,
2529 18206, 18210, 18211, 18217, 18219, 18220, 18221, 18234, 18244, 18245,
2530 18247, 18287, 18319, 18324, 18333, 18346, 18371, 18383, 18397, 18400,
2531 18409, 18410, 18412, 18418, 18422, 18434, 18444, 18457, 18468, 18469,
2532 18470, 18479, 18483, 18495, 18496, 18497, 18498, 18502, 18507, 18508,
2533 18512, 18513, 18519, 18520, 18522, 18527, 18528, 18529, 18530, 18532,
2534 18533, 18534, 18536, 18539, 18540, 18542, 18544, 18545, 18546, 18547,
2535 18549, 18553, 18557, 18558, 18569, 18583, 18585, 18586, 18592, 18593,
2536 18594, 18602, 18612, 18613, 18619, 18633, 18635, 18641, 18643, 18648,
2537 18657, 18676, 18694, 18696, 18887.
2539 * Cache information can be queried via sysconf() function on s390 e.g. with
2540 _SC_LEVEL1_ICACHE_SIZE as argument.
2542 * A buffer overflow in gethostbyname_r and related functions performing DNS
2543 requests has been fixed. If the NSS functions were called with a
2544 misaligned buffer, the buffer length change due to pointer alignment was
2545 not taken into account. This could result in application crashes or,
2546 potentially arbitrary code execution, using crafted, but syntactically
2547 valid DNS responses. (CVE-2015-1781)
2549 * The time zone file parser has been made more robust against crafted time
2550 zone files, avoiding heap buffer overflows related to the processing of
2551 the tzh_ttisstdcnt and tzh_ttisgmtcnt fields, and a stack overflow due to
2552 large time zone data files. Overly long time zone specifiers in the TZ
2553 variable no longer result in stack overflows and crashes.
2555 * A powerpc and powerpc64 optimization for TLS, similar to TLS descriptors
2556 for LD and GD on x86 and x86-64, has been implemented. You will need
2557 binutils-2.24 or later to enable this optimization.
2559 * Character encoding and ctype tables were updated to Unicode 7.0.0, using
2560 new generator scripts contributed by Pravin Satpute and Mike FABIAN (Red
2561 Hat). These updates cause user visible changes, such as the fix for bug
2564 * CVE-2014-8121 The NSS backends shared internal state between the getXXent
2565 and getXXbyYY NSS calls for the same database, causing a denial-of-service
2566 condition in some applications.
2568 * Added vector math library named libmvec with the following vectorized x86_64
2569 implementations: cos, cosf, sin, sinf, sincos, sincosf, log, logf, exp, expf,
2571 The library can be disabled with --disable-mathvec. Use of the functions is
2572 enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0.
2573 Shared library libmvec.so is linked in as needed when using -lm (no need to
2574 specify -lmvec explicitly for not static builds).
2575 Visit <https://sourceware.org/glibc/wiki/libmvec> for detailed information.
2577 * A new fmemopen implementation has been added with the goal of POSIX
2578 compliance. The new implementation fixes the following long-standing
2579 issues: BZ#6544, BZ#11216, BZ#12836, BZ#13151, BZ#13152, and BZ#14292. The
2580 old implementation is still present for use be by existing binaries.
2582 * The 32-bit sparc sigaction ABI was inadvertently broken in the 2.20 and 2.21
2583 releases. It has been fixed to match 2.19 and older, but binaries built
2584 against 2.20 and 2.21 might need to be recompiled. See BZ#18694.
2586 * Port to Native Client running on ARMv7-A (--host=arm-nacl).
2587 Contributed by Roland McGrath (Google).
2589 * The header <regexp.h> is deprecated, and will be removed in a future
2590 release. Use of this header will trigger a deprecation warning.
2591 Application developers should update their code to use <regex.h> instead.
2593 This header was formerly part of SUS, but was deprecated in 1994 and
2594 removed from the standard in 2001. Also, the glibc implementation
2595 leaks memory. See BZ#18681 for more details.
2599 * The following bugs are resolved with this release:
2601 6652, 10672, 12674, 12847, 12926, 13862, 14132, 14138, 14171, 14498,
2602 15215, 15378, 15884, 16009, 16418, 16191, 16469, 16576, 16617, 16618,
2603 16619, 16657, 16740, 16857, 17192, 17266, 17273, 17344, 17363, 17370,
2604 17371, 17411, 17460, 17475, 17485, 17501, 17506, 17508, 17522, 17555,
2605 17570, 17571, 17572, 17573, 17574, 17582, 17583, 17584, 17585, 17589,
2606 17594, 17601, 17608, 17616, 17625, 17630, 17633, 17634, 17635, 17647,
2607 17653, 17657, 17658, 17664, 17665, 17668, 17682, 17702, 17717, 17719,
2608 17722, 17723, 17724, 17725, 17732, 17733, 17744, 17745, 17746, 17747,
2609 17748, 17775, 17777, 17780, 17781, 17782, 17791, 17793, 17796, 17797,
2610 17801, 17803, 17806, 17834, 17844, 17848, 17868, 17869, 17870, 17885,
2613 * CVE-2015-1472 CVE-2015-1473 Under certain conditions wscanf can allocate
2614 too little memory for the to-be-scanned arguments and overflow the
2615 allocated buffer. The implementation now correctly computes the required
2616 buffer size when using malloc, and switches to malloc from alloca as
2619 * A new semaphore algorithm has been implemented in generic C code for all
2620 machines. Previous custom assembly implementations of semaphore were
2621 difficult to reason about or ensure that they were safe. The new version
2622 of semaphore supports machines with 64-bit or 32-bit atomic operations.
2623 The new semaphore algorithm is used by sem_init, sem_open, sem_post,
2624 sem_wait, sem_timedwait, sem_trywait, and sem_getvalue.
2626 * Port to Altera Nios II has been contributed by Mentor Graphics.
2628 * Optimized strcpy, stpcpy, strncpy, stpncpy, strcmp, and strncmp
2629 implementations for powerpc64/powerpc64le.
2630 Implemented by Adhemerval Zanella (IBM).
2632 * Added support for TSX lock elision of pthread mutexes on powerpc32, powerpc64
2633 and powerpc64le. This may improve lock scaling of existing programs on
2634 HTM capable systems. The lock elision code is only enabled with
2635 --enable-lock-elision=yes. Also, the TSX lock elision implementation for
2636 powerpc will issue a transaction abort on every syscall to avoid side
2637 effects being visible outside transactions.
2639 * Optimized strcpy, stpcpy, strchrnul and strrchr implementations for
2640 AArch64. Contributed by ARM Ltd.
2642 * i386 memcpy functions optimized with SSE2 unaligned load/store.
2644 * CVE-2104-7817 The wordexp function could ignore the WRDE_NOCMD flag
2645 under certain input conditions resulting in the execution of a shell for
2646 command substitution when the application did not request it. The
2647 implementation now checks WRDE_NOCMD immediately before executing the
2648 shell and returns the error WRDE_CMDSUB as expected.
2650 * CVE-2012-3406 printf-style functions could run into a stack overflow when
2651 processing format strings with a large number of format specifiers.
2653 * CVE-2014-9402 The nss_dns implementation of getnetbyname could run into an
2654 infinite loop if the DNS response contained a PTR record of an unexpected
2657 * The minimum GCC version that can be used to build this version of the GNU
2658 C Library is GCC 4.6. Older GCC versions, and non-GNU compilers, can
2659 still be used to compile programs using the GNU C Library.
2661 * The GNU C Library is now built with -Werror by default. This can be
2662 disabled by configuring with --disable-werror.
2664 * New locales: tu_IN, bh_IN, raj_IN, ce_RU.
2666 * The obsolete sigvec function has been removed. This was the original
2667 4.2BSD interface that inspired the POSIX.1 sigaction interface, which
2668 programs have been using instead for about 25 years. Of course, ABI
2669 compatibility for old binaries using sigvec remains intact.
2671 * Merged gettext 0.19.3 into the intl subdirectory. This fixes building
2672 with newer versions of bison.
2674 * Support for MIPS o32 FPXX, FP64A and FP64 ABI Extensions.
2675 The original MIPS o32 hard-float ABI requires an FPU where double-precision
2676 registers overlay two consecutive single-precision registers. MIPS32R2
2677 introduced a new FPU mode (FR=1) where double-precision registers extend the
2678 corresponding single-precision registers which is incompatible with the
2679 o32 hard-float ABI. The MIPS SIMD ASE and the MIPSR6 architecture both
2680 require the use of FR=1 making a transition necessary. New o32 ABI
2681 extensions enable users to migrate over time from the original o32 ABI
2682 through to the updated o32 FP64 ABI. To achieve this the dynamic linker now
2683 tracks the ABI of any loaded object and verifies that new objects are
2684 compatible. Mode transitions will also be requested as required and
2685 unsupportable objects will be rejected. The ABI checks include both soft and
2686 hard float ABIs for o32, n32 and n64.
2688 GCC 5 with GNU binutils 2.25 onwards:
2689 It is strongly recommended that all o32 system libraries are built using the
2690 new o32 FPXX ABI (-mfpxx) to facilitate the transition as this is compatible
2691 with the original and all new o32 ABI extensions. Configure a MIPS GCC
2692 compiler using --with-fp-32=xx to set this by default.
2696 * The following bugs are resolved with this release:
2698 6804, 9894, 11505, 12994, 13347, 13651, 14308, 14770, 15119, 15132, 15347,
2699 15514, 15698, 15804, 15894, 15946, 16002, 16064, 16095, 16194, 16198,
2700 16275, 16284, 16287, 16315, 16348, 16349, 16354, 16357, 16362, 16447,
2701 16516, 16532, 16539, 16545, 16561, 16562, 16564, 16574, 16599, 16600,
2702 16609, 16610, 16611, 16613, 16619, 16623, 16629, 16632, 16634, 16639,
2703 16642, 16648, 16649, 16670, 16674, 16677, 16680, 16681, 16683, 16689,
2704 16695, 16701, 16706, 16707, 16712, 16713, 16714, 16724, 16731, 16739,
2705 16740, 16743, 16754, 16758, 16759, 16760, 16770, 16786, 16789, 16791,
2706 16796, 16799, 16800, 16815, 16823, 16824, 16831, 16838, 16839, 16849,
2707 16854, 16876, 16877, 16878, 16882, 16885, 16888, 16890, 16892, 16912,
2708 16915, 16916, 16917, 16918, 16922, 16927, 16928, 16932, 16943, 16958,
2709 16965, 16966, 16967, 16977, 16978, 16984, 16990, 16996, 17009, 17022,
2710 17031, 17042, 17048, 17050, 17058, 17061, 17062, 17069, 17075, 17078,
2711 17079, 17084, 17086, 17088, 17092, 17097, 17125, 17135, 17137, 17150,
2712 17153, 17187, 17213, 17259, 17261, 17262, 17263, 17319, 17325, 17354.
2714 * Reverted change of ABI data structures for s390 and s390x:
2715 On s390 and s390x the size of struct ucontext and jmp_buf was increased in
2716 2.19. This change is reverted in 2.20. The introduced 2.19 symbol versions
2717 of getcontext, setjmp, _setjmp, __sigsetjmp, longjmp, _longjmp, siglongjmp
2718 are preserved pointing straight to the same implementation as the old ones.
2719 Given that, new callers will simply provide a too-big buffer to these
2720 functions. Any applications/libraries out there that embed jmp_buf or
2721 ucontext_t in an ABI-relevant data structure that have already been rebuilt
2722 against 2.19 headers will have to rebuilt again. This is necessary in any
2723 case to revert the breakage in their ABI caused by the glibc change.
2725 * Support for file description locks is added to systems running the
2726 Linux kernel. The standard file locking interfaces are extended to
2727 operate on file descriptions, not file descriptors, via the use of
2728 F_OFD_GETLK, F_OFD_SETLK, and F_OFD_SETLKW. File description locks
2729 are associated with an open file instead of a process.
2731 * Optimized strchr implementation for AArch64. Contributed by ARM Ltd.
2733 * The minimum Linux kernel version that this version of the GNU C Library
2734 can be used with is 2.6.32.
2736 * Running the testsuite no longer terminates as soon as a test fails.
2737 Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
2738 with PASS or FAIL lines for individual tests. A summary of the results is
2739 printed, including a list of failing lists, and "make check" exits with
2740 error status if there were any unexpected failures. "make check
2741 stop-on-test-failure=y" may be used to keep the old behavior.
2743 * The am33 port, which had not worked for several years, has been removed
2746 * The _BSD_SOURCE and _SVID_SOURCE feature test macros are no longer
2747 supported; they now act the same as _DEFAULT_SOURCE (but generate a
2748 warning). Except for cases where _BSD_SOURCE enabled BSD interfaces that
2749 conflicted with POSIX (support for which was removed in 2.19), the
2750 interfaces those macros enabled remain available when compiling with
2751 _GNU_SOURCE defined, with _DEFAULT_SOURCE defined, or without any feature
2752 test macros defined.
2754 * Optimized strcmp implementation for ARMv7. Contributed by ARM Ltd.
2756 * Added support for TX lock elision of pthread mutexes on s390 and s390x.
2757 This may improve lock scaling of existing programs on TX capable systems.
2758 The lock elision code is only built with --enable-lock-elision=yes and
2759 then requires a GCC version supporting the TX builtins. With lock elision
2760 default mutexes are elided via __builtin_tbegin, if the cpu supports
2761 transactions. By default lock elision is not enabled and the elision code
2764 * CVE-2014-4043 The posix_spawn_file_actions_addopen implementation did not
2765 copy the path argument. This allowed programs to cause posix_spawn to
2766 deference a dangling pointer, or use an unexpected pathname argument if
2767 the string was modified after the posix_spawn_file_actions_addopen
2770 * All supported architectures now use the main glibc sysdeps directory
2771 instead of some being in a separate "ports" directory (which was
2772 distributed separately before glibc 2.17).
2774 * The NPTL implementation of POSIX pthreads is no longer an "add-on".
2775 On configurations that support it (all Linux configurations), it's now
2776 used regardless of the --enable-add-ons switch to configure. It is no
2777 longer possible to build such configurations without pthreads support.
2779 * Locale names, including those obtained from environment variables (LANG
2780 and the LC_* variables), are more tightly checked for proper syntax.
2781 setlocale will now fail (with EINVAL) for locale names that are overly
2782 long, contain slashes without starting with a slash, or contain ".." path
2783 components. (CVE-2014-0475) Previously, some valid locale names were
2784 silently replaced with the "C" locale when running in AT_SECURE mode
2785 (e.g., in a SUID program). This is no longer necessary because of the
2788 * On x86-64, the dynamic linker's lazy-binding support is now compatible
2789 with application code using Intel MPX instructions. (With all previous
2790 versions, the MPX register state could be clobbered when making calls
2791 into or out of a shared library.) Note that while the new dynamic
2792 linker is compatible with all known x86 hardware whether or not it
2793 supports Intel MPX, some x86 instruction-set emulators might fail to
2794 handle the new instruction encodings. This is known to affect Valgrind
2795 versions up through 3.9 (but will be fixed in the forthcoming 3.10
2796 release), and might affect other tools that do instruction emulation.
2798 * Support for loadable gconv transliteration modules has been removed.
2799 The support for transliteration modules has been non-functional for
2800 over a decade, and the removal is prompted by security defects. The
2801 normal gconv conversion modules are still supported. Transliteration
2802 with //TRANSLIT is still possible, and the //IGNORE specifier
2803 continues to be supported. (CVE-2014-5119)
2805 * Decoding a crafted input sequence in the character sets IBM933, IBM935,
2806 IBM937, IBM939, IBM1364 could result in an out-of-bounds array read,
2807 resulting a denial-of-service security vulnerability in applications which
2808 use functions related to iconv. (CVE-2014-6040)
2812 * The following bugs are resolved with this release:
2814 156, 387, 431, 762, 832, 926, 2801, 4772, 6786, 6787, 6807, 6810, 6981,
2815 7003, 9721, 9954, 10253, 10278, 11087, 11157, 11214, 12100, 12486, 12751,
2816 12986, 13028, 13982, 13985, 14029, 14032, 14120, 14143, 14155, 14286,
2817 14547, 14699, 14752, 14782, 14876, 14910, 15004, 15048, 15073, 15089,
2818 15128, 15218, 15268, 15277, 15308, 15362, 15374, 15400, 15425, 15427,
2819 15483, 15522, 15531, 15532, 15593, 15601, 15608, 15609, 15610, 15632,
2820 15640, 15670, 15672, 15680, 15681, 15723, 15734, 15735, 15736, 15748,
2821 15749, 15754, 15760, 15763, 15764, 15797, 15799, 15825, 15843, 15844,
2822 15846, 15847, 15849, 15850, 15855, 15856, 15857, 15859, 15867, 15886,
2823 15887, 15890, 15892, 15893, 15895, 15897, 15901, 15905, 15909, 15915,
2824 15917, 15919, 15921, 15923, 15939, 15941, 15948, 15963, 15966, 15968,
2825 15985, 15988, 15997, 16032, 16034, 16036, 16037, 16038, 16041, 16046,
2826 16055, 16071, 16072, 16074, 16077, 16078, 16103, 16112, 16143, 16144,
2827 16146, 16150, 16151, 16153, 16167, 16169, 16172, 16195, 16214, 16245,
2828 16271, 16274, 16283, 16289, 16293, 16314, 16316, 16330, 16337, 16338,
2829 16356, 16365, 16366, 16369, 16372, 16375, 16379, 16381, 16384, 16385,
2830 16386, 16387, 16390, 16394, 16398, 16400, 16407, 16408, 16414, 16430,
2831 16431, 16453, 16474, 16506, 16510, 16529
2833 * Slovenian translations for glibc messages have been contributed by the
2834 Translation Project's Slovenian team of translators.
2836 * The public headers no longer use __unused nor __block. This change is to
2837 support compiling programs that are derived from BSD sources and use
2838 __unused internally, and to support compiling with Clang's -fblock
2839 extension which uses __block.
2841 * CVE-2012-4412 The strcoll implementation caches indices and rules for
2842 large collation sequences to optimize multiple passes. This cache
2843 computation may overflow for large collation sequences and may cause a
2844 stack or buffer overflow. This is now fixed to use a slower algorithm
2845 which does not use a cache if there is an integer overflow.
2847 * CVE-2012-4424 The strcoll implementation uses malloc to cache indices and
2848 rules for large collation sequences to optimize multiple passes and falls
2849 back to alloca if malloc fails, resulting in a possible stack overflow.
2850 The implementation now falls back to an uncached collation sequence lookup
2853 * CVE-2013-4788 The pointer guard used for pointer mangling was not
2854 initialized for static applications resulting in the security feature
2855 being disabled. The pointer guard is now correctly initialized to a
2856 random value for static applications. Existing static applications need
2857 to be recompiled to take advantage of the fix (bug 15754).
2859 * CVE-2013-4237 The readdir_r function could write more than NAME_MAX bytes
2860 to the d_name member of struct dirent, or omit the terminating NUL
2861 character. (Bugzilla #14699).
2863 * CVE-2013-4332 The pvalloc, valloc, memalign, posix_memalign and
2864 aligned_alloc functions could allocate too few bytes or corrupt the
2865 heap when passed very large allocation size values (Bugzilla #15855,
2868 * CVE-2013-4458 Stack overflow in getaddrinfo with large number of results
2869 for AF_INET6 has been fixed (Bugzilla #16072).
2871 * New locales: ak_GH, anp_IN, ar_SS, cmn_TW, hak_TW, lzh_TW, nan_TW, pap_AW,
2872 pap_CW, quz_PE, the_NP.
2874 * Substantially revised locales: gd_GB, ht_HT
2876 * The LC_ADDRESS field was updated to support country_car for almost all
2879 * ISO 1427 definitions were updated.
2881 * ISO 3166 definitions were updated.
2883 * The localedef utility now supports --big-endian and --little-endian
2884 command-line options to generate locales for a different system from that
2885 for which the C library was built.
2887 * Binary locale files now only depend on the endianness of the system for
2888 which they are generated and not on other properties of that system. As a
2889 consequence, binary files generated with new localedef may be incompatible
2890 with old versions of the GNU C Library, and binary files generated with
2891 old localedef may be incompatible with this version of the GNU C Library,
2892 in the following circumstances:
2894 + Locale files may be incompatible on m68k systems.
2896 + Locale archive files (but not separate files for individual locales) may
2897 be incompatible on systems where plain "char" is signed.
2899 * The configure option --disable-versioning has been removed. Builds with
2900 --disable-versioning had not worked for several years.
2902 * ISO 639 definitions were updated for Chiga (cgg) and Chinese (gan, hak, czh,
2903 cjy, lzh, cmn, mnp, cdo, czo, cpx, wuu, hsn, yue).
2905 * SystemTap probes for malloc have been introduced.
2907 * SystemTap probes for slow multiple precision fallback paths of
2908 transcendental functions have been introduced.
2910 * Support for powerpc64le has been added.
2912 * The soft-float powerpc port now supports e500 processors.
2914 * Support for STT_GNU_IFUNC symbols added for ppc32/power4+ and ppc64.
2916 * A new feature test macro _DEFAULT_SOURCE is available to enable the same
2917 set of header declarations that are enabled by default, even when other
2918 feature test macros or compiler options such as -std=c99 would otherwise
2919 disable some of those declarations.
2921 * The _BSD_SOURCE feature test macro no longer enables BSD interfaces that
2922 conflict with POSIX. The libbsd-compat library (which was a dummy library
2923 that did nothing) has also been removed.
2925 * Preliminary documentation about Multi-Thread, Async-Signal and
2926 Async-Cancel Safety has been added.
2928 * Change of ABI data structures for s390 and s390x:
2929 On s390 and s390x the size of struct ucontext and jmp_buf was increased to
2930 allow for future hardware extensions. All library functions that accept or
2931 return these structures were versioned in order to provide backwards
2932 compatibility with existing code. However, not all cases can be handled
2933 with interface versioning. If an application embeds either structure into
2934 another structure and that structure is passed to another compilation unit
2935 which expects the newer larger-sized object then any access to the new
2936 fields will result in undefined behaviour. Similarly any access to fields
2937 that were moved by the enlarging of the structures will cause undefined
2938 behaviour. To fix the undefined behaviour all objects that are part of the
2939 final in-memory image must agree on the size of structures, and this may
2940 require recompilation.
2944 * The following bugs are resolved with this release:
2946 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120,
2947 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988,
2948 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496,
2949 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14941,
2950 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003,
2951 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062,
2952 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234,
2953 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335,
2954 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381,
2955 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424,
2956 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485,
2957 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577,
2958 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711,
2961 * CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal
2962 has been fixed by disabling the use of pt_chown (Bugzilla #15755).
2963 Distributions can re-enable building and using pt_chown via the new configure
2964 option `--enable-pt_chown'. Enabling the use of pt_chown carries with it
2965 considerable security risks and should only be used if the distribution
2966 understands and accepts the risks.
2968 * CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla
2971 * CVE-2013-1914 Stack overflow in getaddrinfo with many results has been
2972 fixed (Bugzilla #15330).
2974 * Add support for calling C++11 thread_local object destructors on thread
2975 and program exit. This needs compiler support for offloading C++11
2976 destructor calls to glibc.
2978 * Improved worst case performance of libm functions with double inputs and
2981 * Support for priority inherited mutexes in pthread condition variables on
2982 non-x86 architectures.
2984 * Port to Xilinx MicroBlaze contributed by David Holsgrove.
2986 * Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
2988 * Optimized string functions for ARM. Implemented by Will Newton and
2991 * Optimized libm functions for SPARC. Implemented by David S. Miller.
2993 * Support for building more of ARM glibc as Thumb-2 code. Implemented by
2996 * Support for building most of MIPS glibc as MIPS16 code. Implemented by
2997 Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
2999 * Added a benchmark framework to track performance of functions in glibc.
3001 * New <math.h> macro named issignaling to check for a signaling NaN (sNaN).
3002 It is based on draft TS 18661 and currently enabled as a GNU extension.
3004 * On Linux, the clock function now uses the clock_gettime system call
3005 for improved precision, rather than old times system call.
3007 * Added support for version-3 tz format files. This is needed when using
3008 the tz database release 2013e or later, and affects a few unusual cases --
3009 currently only TZ='America/Godthab' for time stamps after 2037.
3011 * Added new API functions pthread_getattr_default_np and
3012 pthread_setattr_default_np to get and set the default pthread
3013 attributes of a process.
3015 * Added support for TSX lock elision for pthread mutexes on i386 and x86-64.
3016 This may improve lock scaling of existing programs on TSX capable systems.
3017 When the --enable-lock-elision=yes parameter is specified at configure
3018 time lock elision will be enabled by default for all PTHREAD_MUTEX_DEFAULT
3021 * Support added for AT_HWCAP2 (to coincide with Linux kernel 3.10
3022 availability). Implemented by Ryan S. Arnold.
3024 * Support added for POWER8 platform. Implemented by Ryan S. Arnold.
3028 * The following bugs are resolved with this release:
3030 1349, 3439, 3479, 3665, 5044, 5246, 5298, 5400, 6530, 6677, 6778, 6808,
3031 9685, 9914, 10014, 10038, 10114, 10191, 10631, 10873, 11438, 11607, 11638,
3032 11741, 12140, 13013, 13412, 13542, 13601, 13603, 13604, 13629, 13679,
3033 13696, 13698, 13717, 13741, 13759, 13761, 13763, 13881, 13939, 13950,
3034 13952, 13966, 14042, 14047, 14090, 14150, 14151, 14152, 14154, 14157,
3035 14166, 14173, 14195, 14197, 14237, 14246, 14251, 14252, 14283, 14298,
3036 14303, 14307, 14328, 14331, 14336, 14337, 14347, 14349, 14368, 14376,
3037 14417, 14447, 14459, 14476, 14477, 14501, 14505, 14510, 14516, 14518,
3038 14519, 14530, 14532, 14538, 14543, 14544, 14545, 14557, 14562, 14568,
3039 14576, 14579, 14583, 14587, 14595, 14602, 14610, 14621, 14638, 14645,
3040 14648, 14652, 14660, 14661, 14669, 14672, 14683, 14694, 14716, 14719,
3041 14743, 14767, 14783, 14784, 14785, 14793, 14796, 14797, 14801, 14803,
3042 14805, 14807, 14811, 14815, 14821, 14822, 14824, 14828, 14831, 14833,
3043 14835, 14838, 14856, 14863, 14865, 14866, 14868, 14869, 14871, 14872,
3044 14879, 14889, 14893, 14898, 14914.
3046 * Optimization of memcpy for MIPS.
3048 * CVE-2011-4609 svc_run() produces high cpu usage when accept fails with
3049 EMFILE has been fixed (Bugzilla #14889).
3051 * The add-on ports collection is now distributed in the "ports" subdirectory
3052 of the main GNU C Library distribution, rather than separately.
3054 * Port to ARM AArch64 contributed by Linaro.
3056 * Support for STT_GNU_IFUNC symbols added for s390 and s390x.
3057 Optimized versions of memcpy, memset, and memcmp added for System z10 and
3059 Implemented by Andreas Krebbel.
3061 * The new function secure_getenv allows secure access to the environment,
3062 returning NULL if running in a SUID/SGID process. This function replaces
3063 the internal function __secure_getenv.
3065 * SystemTap static probes have been added into the dynamic linker.
3066 Implemented by Gary Benson.
3068 * Optimizations of string functions strstr, strcasestr and memmem.
3069 Implemented by Maxim Kuvyrkov.
3071 * The minimum Linux kernel version that this version of the GNU C Library
3072 can be used with is 2.6.16.
3074 * Optimizations of string functions memchr, wcschr, wcscpy, and wcsrchr for
3075 powerpc POWER7. Implemented by Will Schmidt.
3077 * New configure option --disable-nscd builds the C library such that it
3078 never attempts to contact the Name Service Caching Daemon (nscd).
3079 New configure option --disable-build-nscd avoids building nscd itself;
3080 this is the default if --disable-nscd is used.
3082 * Improved support for cross-compilation, including cross-testing and
3083 bootstrap builds without a previously built glibc.
3085 * Several testsuite tests are now able to test multiple IFUNC variants of an
3086 interface, rather than just testing the one that would be chooen by
3089 * New configure options --with-bugurl and --with-pkgversion, for
3090 distributors to use to embed their bug-reporting and package version
3091 information in --help and --version output.
3093 * The ttyname and ttyname_r functions on Linux now fall back to searching for
3094 the tty file descriptor in /dev/pts or /dev if /proc is not available. This
3095 allows creation of chroots without the procfs mounted on /proc.
3097 * The `crypt' function now fails if passed salt bytes that violate the
3098 specification for those values. On Linux, the `crypt' function will
3099 consult /proc/sys/crypto/fips_enabled to determine if "FIPS mode" is
3100 enabled, and fail on encrypted strings using the MD5 or DES algorithm
3101 when the mode is enabled.
3103 * The `clock_*' suite of functions (declared in <time.h>) is now available
3104 directly in the main C library. Previously it was necessary to link with
3105 -lrt to use these functions. This change has the effect that a
3106 single-threaded program that uses a function such as `clock_gettime' (and
3107 is not linked with -lrt) will no longer implicitly load the pthreads
3108 library at runtime and so will not suffer the overheads associated with
3109 multi-thread support in other code such as the C++ runtime library.
3111 * New locales: ayc_PE, doi_IN, ia_FR, mni_IN, nhn_MX, niu_NU, niu_NZ,
3116 * The following bugs are resolved with this release:
3118 174, 208, 350, 369, 411, 706, 766, 2074, 2541, 2547, 2548, 2550, 2551,
3119 2552, 2553, 2554, 2562, 2563, 2565, 2566, 2570, 2576, 2636, 2678, 3335,
3120 3440, 3748, 3768, 3866, 3868, 3906, 3976, 3992, 4026, 4108, 4596, 4822,
3121 5077, 5461, 5805, 5993, 6471, 6486, 6578, 6649, 6730, 6770, 6794, 6884,
3122 6890, 6894, 6895, 6907, 6911, 6959, 7064, 9739, 9902, 10110, 10135, 10140,
3123 10153, 10210, 10254, 10346, 10375, 10545, 10716, 10846, 10882, 11174,
3124 11261, 11322, 11365, 11451, 11494, 11521, 11677, 11837, 11959, 12047,
3125 12097, 12193, 12194, 12297, 12298, 12301, 12340, 12354, 12416, 12495,
3126 13058, 13223, 13361, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
3127 13532, 13533, 13547, 13551, 13552, 13553, 13555, 13556, 13559, 13563,
3128 13566, 13576, 13579, 13583, 13592, 13594, 13613, 13618, 13637, 13656,
3129 13658, 13673, 13691, 13695, 13704, 13705, 13706, 13718, 13726, 13738,
3130 13739, 13743, 13750, 13758, 13760, 13761, 13775, 13786, 13787, 13792,
3131 13806, 13824, 13840, 13841, 13844, 13846, 13848, 13851, 13852, 13854,
3132 13871, 13872, 13873, 13879, 13882, 13883, 13884, 13885, 13886, 13892,
3133 13895, 13908, 13910, 13911, 13912, 13913, 13914, 13915, 13916, 13917,
3134 13918, 13919, 13920, 13921, 13922, 13923, 13924, 13926, 13927, 13928,
3135 13938, 13941, 13942, 13954, 13955, 13956, 13963, 13967, 13968, 13970,
3136 13973, 13979, 13983, 13986, 13996, 14012, 14027, 14033, 14034, 14036,
3137 14040, 14043, 14044, 14048, 14049, 14050, 14053, 14055, 14059, 14064,
3138 14075, 14080, 14083, 14103, 14104, 14109, 14112, 14117, 14122, 14123,
3139 14134, 14153, 14183, 14188, 14199, 14210, 14218, 14229, 14241, 14273,
3142 * Support for the x32 ABI on x86-64 added. The x32 target is selected by
3143 configuring glibc with:
3144 BUILD_CC='gcc' CC='gcc -mx32' CXX='g++ -mx32'
3145 Visit <http://sites.google.com/site/x32abi/> for more x32 ABI info.
3146 Implemented by H.J. Lu.
3150 + define static_assert
3152 + do not declare gets
3154 + declare at_quick_exit and quick_exit also for ISO C11
3156 + aligned_alloc. NB: The code is deliberately allows the size parameter
3157 to not be a multiple of the alignment. This is a moronic requirement
3158 in the standard but it is only a requirement on the caller, not the
3161 + timespec_get added
3163 + uchar.h support added
3165 + CMPLX, CMPLXF, CMPLXL added
3167 Implemented by Ulrich Drepper.
3169 * Support for the IA-64 has been moved to ports.
3171 * Remove support for anything but ELF binary format
3173 * Checking versions of poll, ppoll added.
3174 Implemented by Ulrich Drepper.
3176 * More generic and 64-bit performance optimizations to math functions.
3177 Implemented by Ulrich Drepper.
3179 * New configure option --enable-obsolete-rpc makes the deprecated RPC
3180 headers and functions available at compile time as they were before
3181 version 2.14. This option will be removed at some time in the future
3182 after the TI-RPC library becomes fully sufficient for the needs of
3183 existing applications.
3185 * Compatibility code for Linux kernel versions before 2.4 has been removed.
3186 Note that glibc is not expected to work with any Linux kernel version
3189 * New header <sys/auxv.h> and function getauxval allowing easy access to
3190 the AT_* key-value pairs passed from the Linux kernel. The header also
3191 defines the HWCAP_* bits associated with the AT_HWCAP key.
3193 * New locales: mag_IN
3195 * New configure option --enable-systemtap builds SystemTap static probes
3196 into libc for setjmp and longjmp and into libpthread for various operations.
3197 So far the setjmp/longjmp probes and some of the libpthread probes are
3198 provided only for i*86 and x86_64.
3199 Implemented by Roland McGrath and Rayson Ho.
3201 * Optimized expf for x86-32 and x86-64. Implemented by Liubov Dmitrieva.
3203 * More optimized functions for PowerPC. Implemented by Adhemerval Zanella
3206 * More optimized functions for SPARC. Implemented by David S. Miller.
3208 * Improved support for cross-compilation, especially bootstrap builds
3209 without a previously built glibc.
3211 * Ports for the TILE-Gx and TILEPro families of processors. Contributed by
3212 Chris Metcalf from Tilera.
3214 * Support for the old ARM ABI has been removed from ports. Only the EABI is
3215 now supported for ARM processors.
3217 * The hard-float variant of the ARM EABI now uses /lib/ld-linux-armhf.so.3
3218 as the name of the dynamic linker, to distinguish it from the
3219 /lib/ld-linux.so.3 dynamic linker used for the base EABI.
3221 * Support for CRIS has been removed from ports.
3223 * A new class of installed header has been documented for low-level
3224 platform-specific functionality. PowerPC added the first instance with a
3225 function to provide time base register access. Contributed by Tulio
3226 Magno Quites Machado Filho.
3228 * ix86 configurations now install header files that are consistent with
3229 what x86-64 configurations install. These same header files can be used
3230 for -m32, -m64, or -mx32 builds.
3231 Contributed by H.J. Lu.
3233 * Math library bug fixes. A thorough audit of all open math library bugs was
3234 conducted by Joseph Myers. Significant progress was made on many math
3235 library bugs resulting in more accurate exceptions and function results.
3236 Many thanks to all those that contributed including Andreas Jaeger for his
3237 patch review and work on the x87 trigonometric instruction issues.
3239 * Timezone data is no longer installed. Timezone-related binaries and scripts
3240 will continue to be installed. Users should obtain their timezone data from
3241 their distribution provider or from the tzdata package at
3242 <ftp://munnari.oz.au/pub/>.
3246 * The following bugs are resolved with this release:
3248 6779, 6783, 9696, 10103, 10709, 11589, 11929, 12403, 12786, 12840, 12847,
3249 12868, 12852, 12874, 12885, 12892, 12906, 12907, 12922, 12935, 12962,
3250 13007, 13021, 13061, 13062, 13067, 13068, 13085, 13088, 13090, 13092,
3251 13096, 13114, 13118, 13123, 13134, 13138, 13147, 13150, 13166, 13179,
3252 13185, 13189, 13192, 13268, 13276, 13282, 13291, 13305, 13328, 13335,
3253 13337, 13344, 13358, 13367, 13413, 13416, 13423, 13439, 13446, 13472,
3254 13484, 13506, 13515, 13523, 13524, 13538, 13540
3256 * New program pldd to list loaded object of a process
3257 Implemented by Ulrich Drepper.
3259 * Add nss_db support back to glibc. No more dependency on Berkeley db
3260 and support for initgroups lookups.
3261 Implemented by Ulrich Drepper.
3263 * Optimized strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-32.
3264 Contributed by HJ Lu.
3266 * Improved strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-64.
3267 Contributed by HJ Lu.
3269 * Optimized strcat, strncat on x86-64 and optimized wcscmp, wcslen, strnlen
3270 on x86-32 and x86-64.
3271 Contributed by Liubov Dmitrieva.
3273 * Optimized strchr and strrchr for SSE on x86-32.
3274 Contributed by Liubov Dmitrieva.
3276 * Optimized memchr, memrchr, rawmemchr, memcmp, wmemcmp, wcschr, wcscpy
3277 for x86-64 and x86-32.
3278 Contributed by Liubov Dmitrieva.
3280 * New interfaces: scandirat, scandirat64
3281 Implemented by Ulrich Drepper.
3283 * Checking versions of FD_SET, FD_CLR, and FD_ISSET added.
3284 Implemented by Ulrich Drepper.
3286 * nscd now also caches the netgroup database.
3287 Implemented by Ulrich Drepper.
3289 * Integrate libm with gcc's -ffinite-math-only option.
3290 Implemented by Ulrich Drepper.
3292 * Lots of generic, 64-bit, and x86-64-specific performance optimizations
3293 to math functions. Implemented by Ulrich Drepper.
3295 * Optimized strcasecmp and strncasecmp for AVX on x86-64.
3296 Implemented by Ulrich Drepper.
3298 * New Linux interfaces: process_vm_readv, process_vm_writev
3300 * Optimized strcasecmp and strncasecmp for SSSE3 and SSE4.2 on x86-32.
3301 Implemented by Ulrich Drepper.
3303 * Optimized nearbyint and strcasecmp for PPC.
3304 Implemented by Adhemerval Zanella.
3306 * New locales: bho_IN, unm_US, es_CU, ta_LK
3310 * The following bugs are resolved with this release:
3312 386, 6420, 7101, 9730, 9732, 9809, 10138, 10149, 10157, 11099, 11257,
3313 11258, 11487, 11532, 11558, 11578, 11634, 11653, 11668, 11697, 11724,
3314 11781, 11799, 11820, 11837, 11857, 11884, 11892, 11895, 11901, 11945,
3315 11947, 11952, 11987, 12052, 12083, 12158, 12178, 12200, 12346, 12350,
3316 12393, 12420, 12432, 12445, 12449, 12453, 12454, 12460, 12469, 12489,
3317 12509, 12510, 12511, 12518, 12527, 12541, 12545, 12551, 12582, 12583,
3318 12587, 12597, 12601, 12611, 12625, 12626, 12631, 12650, 12653, 12655,
3319 12660, 12671, 12681, 12684, 12685, 12711, 12713, 12714, 12717, 12723,
3320 12724, 12734, 12738, 12746, 12766, 12775, 12777, 12782, 12788, 12792,
3321 12795, 12811, 12813, 12814, 12841
3323 * The RPC implementation in libc is obsoleted. Old programs keep working
3324 but new programs cannot be linked with the routines in libc anymore.
3325 Programs in need of RPC functionality must be linked against TI-RPC.
3326 The TI-RPC implementation is IPv6 enabled and there are other benefits.
3328 Visible changes of this change include (obviously) the inability to link
3329 programs using RPC functions without referencing the TI-RPC library and the
3330 removal of the RPC headers from the glibc headers.
3331 Implemented by Ulrich Drepper.
3333 * New Linux interfaces: clock_adjtime, name_to_handle_at, open_by_handle_at,
3334 syncfs, setns, sendmmsg
3336 * New locales: os_RU, bem_ZA, en_ZA, ff_SN, sw_KE, sw_TZ, lb_LU, wae_CH,
3337 yue_HK, lij_IT, mhr_RU
3339 * New encodings: CP770, CP771, CP772, CP773, CP774
3341 * New program sotruss to trace calls through PLTs
3342 Implemented by Ulrich Drepper.
3344 * The malloc hook implementation is marked deprecated and will be removed
3345 from the default implementation in the next version. The design never
3346 worked ever since the introduction of threads. Even programs which do
3347 not create threads themselves can use multiple threads created internally.
3351 * The following bugs are resolved with this release:
3353 3268, 6812, 7066, 10085, 10484, 10563, 10851, 11149, 11155, 11577, 11611,
3354 11640, 11655, 11701, 11840, 11856, 11883, 11903, 11904, 11968, 11979,
3355 12005, 12037, 12067, 12077, 12078, 12092, 12093, 12107, 12108, 12113,
3356 12140, 12159, 12167, 12191, 12194, 12201, 12204, 12205, 12207, 12348,
3359 * New Linux interfaces: prlimit, prlimit64, fanotify_init, fanotify_mark
3361 * POWER7 optimizations: memset, memcmp, strncmp
3363 * New optimized string functions for x86-64: strnlen (SSE2),
3364 strcasecmp (SSE2, SSSE3, SSE4.2), strncasecmp (SSE2, SSSE3, SSE4.2)
3365 Implemented by Ulrich Drepper.
3369 * The following bugs are resolved with this release:
3371 3662, 4457, 5553, 10162, 10401, 10414, 10554, 10824, 10864, 10915,
3372 10918, 10936, 10939, 10958, 10968, 10969, 10972, 10992, 11000, 11001,
3373 11007, 11010, 11027, 11039, 11040, 11041, 11043, 11046, 11056, 11070,
3374 11093, 11115, 11125, 11126, 11127, 11134, 11141, 11149, 11183, 11184,
3375 11185, 11186, 11187, 11188, 11189, 11190, 11191, 11192, 11193, 11194,
3376 11200, 11230, 11235, 11242, 11254, 11258, 11271, 11272, 11276, 11279,
3377 11287, 11292, 11319, 11332, 11333, 11387, 11389, 11390, 11394, 11397,
3378 11410, 11438, 11449, 11470, 11471, 11520, 11537, 11538, 11571
3380 * New interfaces: pthread_getname_np, pthread_setname_np
3382 * New Linux interface: recvmmsg
3384 * STT_GNU_IFUNC implemented for Sparc by David Miller.
3386 * The dynamic linker now recognizes supported ABI versions from the
3387 EI_ABIVERSION field in the ELF header.
3388 Implemented by Ulrich Drepper.
3390 * New NIS mode selector ADJUNCT_AS_SHADOW. The passwd.adjunct.byname table
3391 will not be used to fill in password fields in the passwd.byname replies.
3392 Instead it is used to synthesize the shadow.byname table, should it be
3393 missing. This is a useful mode in some installations involving Solaris.
3394 Implemented by Ulrich Drepper.
3396 * New locales: kok_IN, sq_MK, cv_RU
3400 * The following bugs are resolved with this release:
3402 3270, 7094, 9924, 9986, 10011, 10085, 10107, 10159, 10162, 10166, 10183,
3403 10195, 10196, 10200, 10203, 10207, 10211, 10213, 10217, 10221, 10229,
3404 10262, 10286, 10312, 10315, 10319, 10349, 10360, 10391, 10402, 10416,
3405 10418, 10422, 10425, 10446, 10448, 10452, 10540, 10553, 10560, 10564,
3406 10609, 10643, 10692, 10713, 10717, 10730, 10731, 10742, 10780, 10784,
3407 10789, 10817, 10823, 10840, 10847
3409 * New interfaces: execvpe, pthread_sigqueue, mkstemps, mkstemps64, mkostemps,
3411 Implemented by Ulrich Drepper.
3413 * Checking version of longjmp added that fails if an uninitialized stack
3414 frame would be created. Implemented by Ulrich Drepper.
3416 * STT_GNU_IFUNC is now supported in static executables.
3417 Implemented by H.J. Lu.
3419 * STT_GNU_IFUNC implemented for PPC by Alan Modra.
3421 * New optimized string functions for x86-64: strstr, strcasestr, memcmp,
3422 strcspn, strpbrk, strspn, strcpy, stpcpy, strncpy, strcmp (SSE2, SSE4.2),
3423 strncmp (SSE2, SSE4.2), strchr (SSE4.2), strrchr (SSE4.2).
3424 Contributed by H.J. Lu.
3426 strlen, rawmemchr, strcmp (SSSE3), strncmp (SSSE3).
3427 Implemented by Ulrich Drepper.
3429 * New optimized string functions for x86: strlen, strcspn, strspn, strpbrk,
3431 Contributed by H.J. Lu.
3433 * Support for fma instruction in AVX on x86-64.
3434 Implemented by H.J. Lu and Ulrich Drepper.
3436 * AVX support in x86-64 auditing support in ld.so.
3437 Implemented by H.J. Lu.
3439 * STB_GNU_UNIQUE support added.
3440 Implemented by Ulrich Drepper.
3442 * Implement second fallback mode for optimized DNS lookup for even more
3443 broken environments. If two requests from the same port are not
3444 handled correctly close the socket and open a new one before sending
3445 the second request. The 'single-request-reopen' option in /etc/resolv.conf
3446 can be used to select this mode right away, instead of rediscovering the
3447 necessity is every process again.
3448 Implemented by Ulrich Drepper.
3450 * New resolver flag RES_USE_DNSSEC to enable use of verified lookup.
3451 Implemented by Adam Tkac.
3453 * Optimized iconv conversions for S390x.
3454 Implemented by Andreas Krebbel.
3456 * Using condvars with PI mutexes is now more efficient due to kernel
3457 support for requeueing to PI futexes. NPTL support added for x86-64.
3458 Implemented by Ulrich Drepper.
3460 * New locales: ps_AF, my_MM
3464 * The following bugs are resolved with this release:
3466 697, 5381, 5807, 6411, 6545, 7040, 7058, 7067, 7080, 7083, 7095, 7098,
3467 9677, 9694, 9697, 9701, 9704, 9705, 9706, 9720, 9726, 9730, 9733, 9736,
3468 9741, 9750, 9753, 9759, 9781, 9791, 9793, 9823, 9833, 9844, 9880, 9881,
3469 9891, 9893, 9895, 9913, 9916, 9920, 9948, 9955, 9957, 9963, 9985, 10052,
3470 10069, 10086, 10087, 10090, 10092, 10093, 10100, 10118, 10128
3472 * New interfaces: preadv, preadv64, pwritev, pwritev64, malloc_info
3473 Implemented by Ulrich Drepper.
3475 * New Linux interfaces: accept4, fallocate, fallocate64.
3476 Implemented by Ulrich Drepper.
3478 * Correct declarations of string function when used in C++ code. This
3479 could lead to compile errors for invalid C++ code.
3481 * XPG7/POSIX 2008 compilation environment. Many GNU-specific functions are
3484 * New POSIX 2008 interface: psiginfo
3485 Implemented by Ulrich Drepper.
3487 * New ISO C++1x interfaces: quick_exit, at_quick_exit
3488 Implemented by Ulrich Drepper.
3490 * Support for selecting between multiple function definitions at runtime
3491 using STT_GNU_IFUNC symbols. Implemented by Ulrich Drepper.
3493 * The libcrypt library can now use the hash function implementations in
3494 NSS. Implemented by Ulrich Drepper.
3496 * The malloc implementation can be compiled to be less memory efficient
3497 but higher performing in multi-threaded programs.
3498 Implemented by Ulrich Drepper.
3500 * New locales: nan_TW@latin, ks_IN
3502 * Faster strlen, strchr, strchrnul, strrchr, memchr, and rawmemchr for x86-64.
3503 Implemented by Ulrich Drepper.
3505 * Extended printf hook support. It is possible to use user-defined types
3506 and extend existing format specifiers.
3507 Implemented by Ulrich Drepper.
3509 * Handling for group shadow files has been added.
3510 Implemented by Ulrich Drepper.
3512 * DNS IPv4-IPv6 parallel lookup now deals better with broken DNS
3513 servers (the case, e.g., for some people using the built-in DNS
3514 server in ADSL modems/routers). There is a once-per-process timeout
3515 in case of a broken server. To avoid it, users can run nscd or put
3516 'options single-request' in /etc/resolv.conf.
3517 Implemented by Ulrich Drepper.
3521 * The following bugs are resolved with this release:
3523 3406, 5209, 5210, 5381, 5794, 5814, 5911, 6428, 6438, 6442, 6447, 6461,
3524 6472, 6505, 6544, 6589, 6612, 6634, 6653, 6654, 6657, 6698, 6712, 6713,
3525 6719, 6723, 6724, 6740, 6763, 6771, 6790, 6791, 6817, 6824, 6839, 6843,
3526 6845, 6856, 6857, 6860, 6867, 6875, 6919, 6920, 6942, 6947, 6955, 6966,
3527 6968, 6974, 6980, 6995, 7008, 7009, 7029
3529 * Unified lookup for getaddrinfo: IPv4 and IPv6 addresses are now looked
3530 up at the same time. Implemented by Ulrich Drepper.
3532 * TLS descriptors for LD and GD on x86 and x86-64.
3533 Implemented by Alexandre Oliva.
3535 * getaddrinfo now handles DCCP and UDPlite.
3536 Implemented by Ulrich Drepper.
3538 * New fixed-size conversion macros: htobe16, htole16, be16toh, le16toh,
3539 htobe32, htole32, be32toh, le32toh, htobe64, htole64, be64toh, le64toh.
3540 Implemented by Ulrich Drepper.
3542 * New implementation of memmem, strstr, and strcasestr which is O(n).
3543 Implemented by Eric Blake.
3545 * New Linux interfaces: inotify_init1, dup3, epoll_create1, pipe2
3547 * Implement "e" option for popen to open file descriptor with the
3548 close-on-exec flag set. Implemented by Ulrich Drepper.
3550 * Implement "b" mode for fmemopen. In this mode writes writes don't
3551 implicitly add a NUL byte and seeks from the end of the buffer really
3552 use the buffer end, not the string length as the basis.
3553 Implemented by Ulrich Drepper.
3555 * Many functions, exported and internal, now atomically set the close-on-exec
3556 flag when run on a sufficiently new kernel. Implemented by Ulrich Drepper.
3558 * Sorting rules for some Indian languages (Kannada, Gurumukhi, Telugu,
3560 Implemented by Pravin Satpute.
3562 * New locales: sd_IN, sd_IN@devanagari, ks_IN@devanagari, ht_HT, en_AG, nl_AW.
3566 * The following bugs are resolved with this release:
3568 2549, 3112, 4314, 4407, 4997, 5012, 5040, 5112, 5204, 5208, 5209, 5220,
3569 5222, 5224, 5225, 5237, 5238, 5240, 5245, 5277, 5346, 5375, 5378, 5382,
3570 5424, 5427, 5428, 5435, 5436, 5439, 5441, 5442, 5443, 5451, 5452, 5454,
3571 5455, 5463, 5464, 5465, 5475, 5477, 5478, 5520, 5541, 5545, 5553, 5600,
3572 5602, 5607, 5614, 5627, 5628, 5736, 5737, 5741, 5753, 5760, 5762, 5768,
3573 5774, 5777, 5778, 5779, 5786, 5790, 5818, 5854, 5857, 5903, 5939, 5979,
3574 5983, 5995, 5998, 6004, 6007, 6020, 6021, 6024, 6040, 6042
3576 * New locales: bo_CN, bo_IN, shs_CA.
3578 * New encoding: HP-ROMAN9, HP-GREEK8, HP-THAI8, HP-TURKISH8.
3580 * Sorting rules for some Indian languages (Devanagari and Gujarati).
3581 Implemented by Pravin Satpute.
3583 * IPV6 addresses in /etc/resolv.conf can now have a scope ID
3585 * nscd caches now all timeouts for DNS entries
3586 Implemented by Ulrich Drepper.
3588 * nscd is more efficient and wakes up less often.
3589 Implemented by Ulrich Drepper.
3591 * More checking functions: asprintf, dprintf, obstack_printf, vasprintf,
3592 vdprintf, and obstack_vprintf.
3593 Implemented by Jakub Jelinek.
3595 * Faster memset for x86-64.
3596 Implemented by Harsha Jagasia and H.J. Lu.
3598 * Faster memcpy on x86.
3599 Implemented by Ulrich Drepper.
3601 * ARG_MAX is not anymore constant on Linux. Use sysconf(_SC_ARG_MAX).
3602 Implemented by Ulrich Drepper.
3604 * Faster sqrt and sqrtf implemention for some PPC variants.
3605 Implemented by Steven Munroe.
3609 * The following bugs are resolved with this release:
3611 4125, 4126, 4178, 4359, 4407, 4512, 4514, 4525, 4554, 4556, 4557, 4566,
3612 4582, 4586, 4588, 4599, 4610, 4647, 4702, 4705, 4726, 4745, 4772, 4773,
3613 4775, 4776, 4792, 4813, 4814, 4816, 4833, 4858, 4860, 4896, 4905, 4925,
3614 4936, 4937, 4938, 4941, 4946, 4963, 4972, 5010, 5028, 5043, 5058, 5063,
3615 5071, 5103, 5104, 5112, 5113, 5184, 5186
3617 * More checking functions: fread, fread_unlocked, open*, mq_open.
3618 Implemented by Jakub Jelinek and Ulrich Drepper.
3620 * Extend fortification to C++. Implemented by Jakub Jelinek.
3622 * Implement 'm' modifier for scanf. Add stricter C99/SUS compliance
3623 by not recognizing 'a' as a modifier when those specs are requested.
3624 Implemented by Jakub Jelinek.
3626 * PPC optimizations to math and string functions.
3627 Implemented by Steven Munroe.
3629 * New interfaces: mkostemp, mkostemp64. Like mkstemp* but allow additional
3630 options to be passed. Implemented by Ulrich Drepper.
3632 * More CPU set manipulation functions. Implemented by Ulrich Drepper.
3634 * New Linux interfaces: signalfd, eventfd, eventfd_read, and eventfd_write.
3635 Implemented by Ulrich Drepper.
3637 * Handle private futexes in the NPTL implementation.
3638 Implemented by Jakub Jelinek and Ulrich Drepper.
3640 * Add support for O_CLOEXEC. Implement in Hurd. Use throughout libc.
3641 Implemented by Roland McGrath and Ulrich Drepper.
3643 * Linux/x86-64 vDSO support. Implemented by Ulrich Drepper.
3645 * SHA-256 and SHA-512 based password encryption.
3646 Implemented by Ulrich Drepper.
3648 * New locales: ber_DZ, ber_MA, en_NG, fil_PH, fur_IT, fy_DE, ha_NG, ig_NG,
3649 ik_CA, iu_CA, li_BE, li_NL, nds_DE, nds_NL, pap_AN, sc_IT, tk_TM, ug_CN,
3652 + New iconv modules: MAC-CENTRALEUROPE, ISO-8859-9E, KOI8-RU.
3653 Implemented by Ulrich Drepper.
3657 * The following bugs are resolved with this release:
3659 3156, 3213, 3285, 3291, 3306, 3313, 3320, 3322, 3325, 3326, 3334, 3348,
3660 3352, 3363, 3369, 3426, 3427, 3429, 3451, 3458, 3465, 3480, 3483, 3493,
3661 3514, 3515, 3559, 3632, 3664, 3673, 3674, 3745, 3747, 3818, 3840, 3842,
3662 3851, 3855, 3884, 3885, 3902, 3905, 3919, 3944, 3954, 3955, 3957, 3991,
3663 3995, 3996, 4024, 4040, 4069, 4070, 4074, 4076, 4096, 4101, 4102, 4114,
3664 4130, 4131, 4181, 4276, 4306, 4342, 4344, 4349, 4364, 4368, 4381, 4392,
3665 4403, 4405, 4406, 4411, 4438, 4439, 4455, 4456, 4465, 4512, 4514, 4586,
3668 * New Linux interfaces: epoll_pwait, sched_getcpu.
3670 * New generic interfaces: strerror_l.
3672 * nscd can now cache the services database. Implemented by Ulrich Drepper.
3676 * The following bugs are resolved with this release:
3678 39, 68, 192, 352, 388, 531, 935, 1006, 1201, 1203, 1386, 1782, 1783, 1784,
3679 1785, 1786, 1787, 1951, 1974, 1996, 2072, 2098, 2125, 2134, 2158, 2182,
3680 2349, 2376, 2386, 2389, 2415, 2418, 2420, 2423, 2450, 2451, 2466, 2472,
3681 2473, 2477, 2498, 2499, 2501, 2502, 2505, 2507, 2508, 2509, 2517, 2526,
3682 2569, 2571, 2592, 2611, 2625, 2632, 2644, 2662, 2680, 2683, 2684, 2693,
3683 2695, 2703, 2734, 2751, 2764, 2766, 2775, 2792, 2821, 2823, 2832, 2841,
3684 2843, 2883, 2892, 2908, 2914, 2926, 2961, 2978, 2980, 2987, 2997, 2998,
3685 3013, 3018, 3034, 3040, 3044, 3123, 3124, 3137, 3138, 3143, 3155, 3189,
3686 3225, 3251, 3252, 3253, 3273
3688 * For Linux, the sorting of addresses returned by getaddrinfo now also
3689 handles rules 3, 4, and 7 from RFC 3484. I.e., all rules are handled.
3690 Implemented by Ulrich Drepper.
3692 * Allow system admin to configure getaddrinfo with the /etc/gai.conf file.
3693 Implemented by Ulrich Drepper.
3695 * New Linux interfaces: splice, tee, sync_file_range, vmsplice.
3697 * New iconv module for MIK. Contributed by Alexander Shopov.
3699 * For sites with broken group and/or passwd database, the auto-propagate
3700 option of nscd can prevent creating ID lookup entries from the results
3701 of a name lookup and vice versa. This usually is no problem but some
3702 site might have problems with the default behavior.
3703 Implemented by Ulrich Drepper.
3705 * Iterating over entire database in NIS can be slow. With the
3706 SETENT_BATCH_READ option in /etc/default/nss a system admin can decide
3707 to trade time for memory. The entire database will be read at once.
3708 Implemented by Ulrich Drepper.
3710 * The interfaces introduced in RFC 3542 have been implemented by
3713 * Support for the new ELF hash table format was added by Ulrich Drepper.
3715 * Support for priority inheritance mutexes added by Jakub Jelinek and
3718 * Support for priority protected mutexes added by Jakub Jelinek.
3720 * New locales: nr_ZA, pa_PK, ca_AD, ca_FR, ca_IT, el_CY, tr_CY, as_IN,
3721 or_IN, csb_PL, fy_NL, sr_ME.
3725 * More overflow detection functions.
3727 * New iconv converters for IBM1025, IBM1122, IBM1137, IBM1153,
3728 IBM1154, IBM1155, IBM1156, IBM1157, and IBM1158 contributed by Jiro Sekiba.
3730 More converters for IBM803, IBM901, IBM902, IBM921, IBM1008, IBM1097,
3731 IBM1112, IBM1123, IBM1130, IBM1140, IBM1141, IBM1142, IBM1143, IBM1144,
3732 IBM1145, IBM1146, IBM1147, IBM1148, IBM1149, IBM1166, IBM1167, IBM4517,
3733 IBM4899, IBM4909, IBM4971, IBM5347, IBM9030, IBM9066, IBM9448, IBM12712,
3734 IBM16804, IBM1364, IBM1371, IBM1388, IBM1390, and IBM1399 contributed
3735 by Masahide Washizawa.
3737 * It is now possible to install audit libraries for the dynamic linker using
3738 LD_AUDIT. Implemented by Ulrich Drepper.
3740 * The LinuxThreads add-on, providing pthreads on Linux 2.4 kernels, is no
3741 longer supported. The new NPTL implementation requires Linux 2.6 kernels.
3742 For a libc and libpthread that works well on Linux 2.4 kernels, we
3743 recommend using the stable 2.3 branch.
3745 * The new function open_wmemstream acts like open_memstream,
3746 but uses a wchar_t wide character buffer.
3748 * The new function ppoll is an improved version of poll, similar to pselect.
3750 * New interfaces: fdopendir, openat, fstatat, fchmodat, fchownat,
3751 futimesat, faccessat, mkdirat, mkfifoat, mknodat,
3752 renameat, unlinkat, linkat, symlinkat, readlinkat.
3754 * New Linux kernel system calls: unshare,
3755 inotify_init, inotify_add_watch, inotify_rm_watch.
3757 * The euidaccess function is now also known by the alias eaccess,
3758 for compatibility with some other systems.
3760 * Timezone data updated to 2006b version.
3764 * The following bugs are resolved with this release:
3766 38, 253, 549, 622, 653, 721, 758, 851, 877, 915, 934, 955, 961,
3767 1016, 1037, 1076, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086,
3768 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,
3769 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109,
3770 1110, 1111, 1112, 1113, 1125, 1137, 1138, 1249, 1250, 1251, 1252,
3771 1253, 1254, 1350, 1358, 1394, 1438, 1498, 1534
3773 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
3775 * As of this release, GCC 4 can be used to compile the C Library.
3777 * Timezone data updated to 2005m version.
3781 * The following bugs are resolved with this release:
3783 284, 592, 602, 626, 633, 640, 650, 661, 671, 681, 693, 700, 710, 719,
3784 722, 723, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736,
3785 737, 738, 739, 740, 741, 742, 743, 744, 745, 765, 767, 768, 769, 776,
3786 777, 787, 821, 822, 823, 825
3788 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
3792 * Support for RFC 3678. Real implementations exist only for Linux so far.
3793 Implemented by Ulrich Drepper.
3795 * nscd can now cache entries persistently. Expiring entries are reloaded.
3796 For speedups the cache can be shared in memory with client processes.
3797 Implemented by Ulrich Drepper.
3799 * nscd can now perform SELinux checks.
3800 Implemented by Matthew Rickard <mjricka@epoch.ncsc.mil>.
3802 * getaddrinfo queries are now cached. Canonical name lookup is performed
3804 Implemented by Ulrich Drepper.
3806 * The nothrow function attribute is used when headers are used by gcc when
3807 compiling C code. This can avoid significant amounts of exception
3810 * The malloc functions perform more error checking and are stricter when
3811 it comes to reacting on errors. The default action is to terminate
3812 the process after showing an error message. Implemented by Ulrich Drepper.
3814 * Reverse lookups of IPv6 addresses does not use bit string or .ip6.int
3815 lookups anymore unless explicitly requested. Implemented by Ulrich Drepper.
3817 * Namespaces in ld.so are implemented. DSOs can be loaded in separate
3818 namespaces using the new function dlmopen(). This feature is of course,
3819 like most other dynamic loading functionality, not available in statically
3820 linked applications. Implemented by Ulrich Drepper.
3822 * Low-overhead boundary checking variants of string and some stdio functions
3823 were added. These are to be used in conjunction with a gcc patch by
3824 Jakub Jelinek which adds calls to these functions if possible.
3825 Implemented by Jakub Jelinek and Ulrich Drepper.
3827 * Old code for several operating systems and machine architectures that
3828 have not been in working condition in a long time have been removed from
3829 the main source tree maintained by the GNU C Library's maintainers.
3830 These files are now reside in the separate `ports' source module
3831 that is usable as an add-on when building the library.
3835 * New functions `dladdr1' and `dlinfo' in <dlfcn.h> provide more ways to
3836 interrogate the dynamic linker, compatible with the Solaris interface.
3838 * ELF thread-local storage support (TLS) now works on PowerPC and PowerPC64;
3839 implemented by Paul Mackerras, Steven Munroe, and Roland McGrath.
3841 * getifaddrs now uses the netlink interface on Linux to get its information.
3842 Implemented by Thorsten Kukuk.
3844 * getaddrinfo now implements AI_V4MAPPED, AI_ALL, and AI_ADDRCONF.
3845 Implemented by Ulrich Drepper.
3847 * support for non-executable stacks on x86 has been added. Changes mostly
3850 * regex is now much faster for multibyte locales. Changes by Jakub Jelinek
3853 * getaddrinfo now performs destination address selection according to
3858 * Thread-safe interfaces for many functions that access locale data
3859 were added in version 2.3, but these features were omitted from NEWS.
3860 Many functions have variants with an `_l' suffix that take a `locale_t'
3861 object as a parameter rather than consulting the current locale.
3862 The new functions `newlocale', `duplocale', and `freelocale' in <locale.h>
3863 create and maintain `locale_t' objects. Additionally, the new function
3864 `uselocale' sets "the current locale" (as used by functions not so
3865 parameterized) set for an individual thread. These features were added
3866 in version 2.3, implemented by Ulrich Drepper and Roland McGrath.
3868 * The functions getresuid, getresgid, setresuid, and setresgid, which
3869 have long been available on Linux, are now declared in <unistd.h>
3870 and are now also available on the Hurd.
3872 * ELF thread-local storage support (TLS) now works on x86-64.
3874 * The new dynamic string token $LIB is expanded in shared library names.
3875 This normally expands to lib, but on some 64-bit platforms to lib64 instead.
3877 * Aldy Hernandez contributed complete software floating point support for
3878 PowerPC machines with no FPU.
3880 * fexecve is implemented on Linux.
3882 * The `btowc' function should work at least twice as fast due to
3883 specialized callbacks in the iconv modules. Implemented by Bruno Haible.
3885 * With approriate thread add-ons cancelable functions are now implemented
3886 in libc.so as well. No need to call the function in libpthread. This
3887 change allowed to finally disable the incorrect and expensive handling
3888 of weak definition in ld.so.
3890 * Yet more PLT entries in libc.so have been removed. We finally arrived
3891 at the bare minimum. Startup times improved appropriately.
3893 * Support for the new Linux/x86 system call interface was added. The
3894 AT_SYSINFO auxiliary vector entry is recognized and handled.
3898 * Masahide Washizawa contributed iconv modules for IBM1163 and IBM1164
3901 * iconv (the program and the interface) now accepts empty names (excluding
3902 options like //TRANSLIT) to mean "use charset of current locale".
3904 * localedef can now transliterate characters in strings which are not in
3905 the provided charmap. The information from the input locale is used.
3907 * Prelinking support was added for ELF targets. This requires additional
3908 tools and recent versions of the GNU binutils. Contributed by Jakub Jelinek.
3910 * Read-only stdio streams now use mmap to speed up operation by eliminating
3911 copying and buffer underflows. To use add 'm' to the mode string of
3912 the fopen/fdopen/freopen call. Implemented by Ulrich Drepper.
3914 * The malloc functions were completely rewritten by Wolfram Gloger based
3915 on Doug Lea's malloc-2.7.0.c.
3917 * Isamu Hasegawa contributed a completely new and POSIX-conformant
3918 implementation of regex.
3920 * Bruno Haible upgraded the iconv and locale implementation to support
3923 * Contents of the LC_* and LANG environment variables in the CEN style are
3924 not recognized anymore. It never was used. Change by Ulrich Drepper.
3926 * The runtime (ld.so, libc, libpthread for Linux) now can handle the ELF
3927 thread-local storage (TLS) ABI on some platforms.
3928 Changes by Ulrich Drepper. SH support by Kaz Kojima.
3930 * Bruno Haible contributed iconv converters for ISO-2022-JP-3, SHIFT JIS-X0213,
3931 EUC-JISX0213, and TSCII.
3933 * New header <ifaddrs.h> with functions `getifaddrs' and `freeifaddrs':
3934 BSD-compatible interface for getting all network interface addresses.
3935 Implementation for IPv4 by Roland McGrath.
3937 * Loading of locale data is faster due to the introduction of a locale
3938 archive. Implemented by Roland McGrath and Ulrich Drepper.
3940 * Startup times are significantly reduced by not using exported functions
3941 inside the library itself. Changes by Jakub Jelinek, Roland McGrath,
3944 * Steven Munroe contributed a port to PowerPC64/Linux.
3948 * The Hurd now uses the GNU libio implementation of stdio rather than the
3949 old GNU stdio implementation, and uses a new ABI (libc.so.0.3).
3951 * The Hurd on x86 now has the `ioperm' function and <sys/io.h> header file
3952 with the same behavior as the Linux system call of the same name.
3956 * Stephen Moshier implemented log2, log10, powl and cbrtl for the
3957 128-bit long double format.
3959 * Masahide Washizawa contributed iconv modules for IBM1132, IBM1133, IBM1160,
3960 IBM1161, and IBM1162 charsets.
3962 * Andreas Jaeger contributed a port to x86-64/Linux.
3964 * Peter Bruin contributed a port to PowerPC/Hurd.
3966 * libc functions using I/O streams now can handle wide-oriented streams
3969 * optimizations in the dynamic linker. Binaries created by recent binutils
3970 versions start up quicker due to reduced time spent on relocations.
3972 * Support for use of gcc3 added by Jakub Jelinek and HJ Lu.
3976 * Stephen Moshier implemented cosh, expm1, log1p, acos, sinh, tanh,
3977 asinh, atanh, j0 for the 128-bit long double format.
3979 * Bruno Haible updated all the code handling Unicode in some form to
3980 support Unicode 3.1.
3982 * Speed of regex for single-byte locales is back to previous levels.
3983 Patch by Isamu Hasegawa.
3985 * Alpha, SPARC, and IA-64 now also using floating stacks.
3987 * Startup time of internationalized applications greatly improved through
3988 iconv cache. Use iconvconfig to generate the cache file.
3989 Contributed by Ulrich Drepper.
3991 * The IA-64 specific part of ld.so was rewritten to eliminate some pretty
3992 severe performance problems. Patch by David Mosberger.
3994 * The Hurd port got a lot more functionality like AIO, various stdio
3995 extensions, etc. Mainly done by Roland McGrath.
3997 * mtrace can now lookup symbols in shared libraries.
4001 * Intel's IA-64 math library is largely integrated. It provides fast and
4002 accurate implementatations for most basic and standard math functions
4003 in float, double, and long double format.
4005 * Stephen Moshier implemented j0, j1, jn, y0, y1, yn, lgamma, erf, erfc,
4006 and asin for the 96-bit long double format and asin, log, tan for the
4007 128-bit long double format.
4009 * The beginning of a last-bit accurate math library by IBM Haifa were added.
4010 The basic double functions exist today. Contributed by Abraham Ziv
4011 <ziv@il.ibm.com>, Moshe Olshansky <olshansk@il.ibm.com>, Ealan Henis
4012 <ealan@il.ibm.com>, and Anna Reitman <reitman@il.ibm.com>.
4014 * An asynchronous name lookup library was added. The interface is designed
4015 after POSIX AIO. The proposal was circulated beforehand to get comments.
4016 No negative ones came in. Implemented by Ulrich Drepper.
4018 * Port to S390/64bit contributed by Martin Schwidefsky
4019 <schwidefsky@de.ibm.com>.
4021 * David Mosberger <davidm@hpl.hp.com> implemented the setcontext family
4022 of functions for Linux/IA-64.
4024 * The RPC code is now thread safe. Threads can now use the same service
4025 of different services at the same time. Patch by Eric Norum
4026 <eric.norum@usask.ca> with some help by Ulrich Drepper.
4028 * Martin Schwidefsky <schwidefsky@de.ibm.com> implemented the setcontext
4029 family of functions for Linux/S390.
4031 * Ulrich Drepper <drepper@redhat.com> implemented the setcontext family
4032 of functions for Linux/x86.
4034 * Port to Linux/CRIS contributed by Axis Communications.
4038 * Lots of headers were cleaned up. Using the tool in the conform/ subdir
4039 we can now check for namespace violations and missing declarations. The
4040 result is that almost all headers are now Unix-compliant (as defined in
4041 the upcoming XPG6). The negative side is that some programs might need
4042 corrections, too, if they depend on the incorrect form of the headers in
4043 previous versions which defined too many symbols and included too many
4046 * regex now handles multibyte character sets correctly.
4047 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
4049 * iconv (the program) does now conform to the upcoming XPG6 and handles
4050 charmaps. Instead of the charset names the path of charmaps can be
4051 provided and the conversion happens based on this data.
4052 Contributed by Ulrich Drepper.
4054 * The locale program now provides detailed information about the installed
4059 only lists the names of the supported locales
4063 provides details such as country, language, and codeset name.
4064 Contributed by Ulrich Drepper.
4068 * The gencat program now parses the input file according to the charset
4069 selected by the LC_CTYPE category. This is important for stateful
4070 character sets. To make generating catalogs easier there is a way
4071 to overwrite the charset selected by the locale: before the first
4072 message or $ quote line the catalog can contain a line like
4074 $ codeset=ISO-8859-2
4076 to select the charset (ISO-8859-2 in this case).
4078 Implemented by Shinya Hanataka and Ulrich Drepper.
4080 * New codeset conversion modules: IBM-922 (Estonia PC codeset),
4081 IBM-1124 (Ukraine PC codeset), IBM-1129 (Vietnamese PC codeset).
4082 Contributed by Masahide Washizawa <washi@jp.ibm.com>.
4084 * Optimized string functions for Alpha ev6 and ev67 by Richard Henderson
4085 <rth@redhat.com> and Rick Gorton <rick.gorton@alpha-processor.com>.
4087 * The LANGUAGE environment variable is now ignored unless the locale is
4088 changed from the default "C" locale.
4090 * The usual bug fixes.
4094 * Greg McGary added runtime support for bounds checking using gcc's
4095 new -fbounded-pointers option. ix86 target is complete. PowerPC
4098 * Thorsten Kukuk added secure mode support to nscd.
4100 * The Berkeley db libraries have been removed.
4102 Related, the nss_db module is now in a separate package since it
4103 obviously requires a database library being available.
4105 * Wide character I/O streams implemented by Ulrich Drepper.
4107 * Functions from the extended socket API added by Ulrich Drepper.
4109 * Functions feenableexcept and fedisableexcept to control the
4110 behaviour of individual exceptions have been added by Andreas Jaeger.
4112 * ldconfig program added by Andreas Jaeger and Jakub Jelinek.
4114 * The resolver code has been updated from bind 8.2.3-T5B which supports
4115 threads. The integration was done by Andreas Jaeger, Adam D. Bradley,
4118 This change could in some situations effect backward compatibility. Since
4119 now `_res' is a thread-local instead of a global variable, modifying it
4120 in one thread does not have any effect in other threads.
4122 The resolver library was also extended to allow IPv6 as the transport
4123 protocol for the requests. This work was done by Stig Venaas.
4125 * Compatibility code for K&R C compilers has been removed from the
4126 header files. A ISO C compiler is needed to use the library
4127 (conforming to either C89 or C99 standard).
4129 * Complete rewrite of the localedef program to support multibyte character
4130 sets. Implement handling of ISO 14651 and ISO 14652. Rewrite strcoll,
4131 strxfrm, wcscoll, and wcsxfrm functions. Make isw*() functions work.
4132 Implemented by Ulrich Drepper.
4134 Bruno Haible significantly improved the generation and use of the data
4135 structures for the wide character tables.
4137 * Plural handling in gettext implemented by Ulrich Drepper.
4139 * The utmp daemon has been removed.
4141 * The port to MIPS-Linux has been finished by Andreas Jaeger.
4143 * A port to Hitachi SH3 and SH4 has been contributed by Kazumoto Kojima
4146 * POSIX clocks and timers implemented by Kaz Kylheku and Ulrich Drepper.
4148 * POSIX barriers implemented by Kaz Kylheku.
4150 * POSIX spawn function family implemented by Ulrich Drepper.
4152 * POSIX spinlocks are now available.
4154 * Timed wait functions for mutex, rwlock, and semaphores are implemented.
4156 * the configure option --enable-kernel=X.Y.Z allows to strip out
4157 compatibility for kernel versions before X.Y.Z. This is currently only
4158 implemented for Linux.
4160 * the sockaddr_in6 structure changed. The IPv6 working group added a new
4161 field sin6_scope_id. This means that all programs using IPv6 should be
4162 recompiled. Don't expect binary compatibility with previous glibc
4165 * various conversion modules for IBM character sets contributed by
4168 * IA-64 port by Jes Sorensen and HJ Lu.
4182 * New ISO C 9x function _Exit, imaxabs, and imaxdiv are added.
4184 * New xdr functions are added; some rpc functions are now 64bit clean.
4186 * Fixed a number of bugs and memory leaks (especially in NIS+ code).
4188 * Fixed known incompatibilities with glibc 2.0.
4190 * New functions lock64, strchrnul, rawmemchr, getutmp and getutmpx.
4192 * Optimized a number of functions (especially the ELF dynamic loader).
4194 * Update timezone data files.
4196 * lots of charmaps corrections
4198 * some new locale definitions and charmaps
4203 * Richard Henderson corrected size of struct timeval on Linux/Alpha to
4204 conform to POSIX member type requirements. Symbol versions have been
4205 adjusted as needed within the library, and for direct use by applications,
4206 but there is potential for problems if third-party libraries use
4207 struct timeval as part of their interface. This does not present
4208 a problem for X and other "essential" system libraries.
4210 * An additional locale model to support C++ Standard Library locale
4211 model and probably more was implemented by Ulrich Drepper.
4213 * Eric Youngdale and Ulrich Drepper implemented versioning of objects on
4216 * Miles Bader provided the `argp' function family to support hierachical
4217 command line argument parsing, layered on top of getopt.
4219 * strtod accepts new hexadecimal floating-point format from ISO C 9X.
4221 * printf has two new specifiers %a and %A to print hexadecimal flaoting-point
4224 * scanf recognizes the %a and %A format for scanning floating point numbers.
4226 * the new headers <stdint.h> and <inttypes.h> from ISO C 9X provides
4227 information and interfaces for the available integer types.
4229 * about 130 new math functions were added to implement the ISO C9x math
4232 * the new header <complex.h> contains definitions of the complex math
4233 functions from ISO C 9X.
4235 * the new header <tgmath.h> defines generic macros to use complex or
4236 real valued functions.
4238 * Thorsten Kukuk provided an implementation for NIS+, securelevel 0, 1 and 2.
4240 * Andreas Jaeger provided a test suite for the math library.
4242 * Mark Kettenis implemented the utmpx interface and an utmp daemon.
4244 * Ulrich Drepper added character set conversion functions (iconv).
4246 * Optimized string functions have been added.
4248 * The localedata addon is now part of glibc.
4250 * An implementation of profiling shared libraries was added by Ulrich Drepper.
4252 * Thorsten Kukuk and Ulrich Drepper provided an implementation for a caching
4253 daemon for NSS (nscd).
4255 Missing a better place here are some numbers on improvements. Under
4256 Linux 2.1.125 un-tar-ing the kernel sources takes
4260 using local files 12.19s 6.88s 22.91s
4262 using NIS 13.92s 8.91s 26.34s
4264 using NIS & nscd 10.37s 7.34s 25.30s
4266 using NIS+ 27.57s 30.37s 640.46s
4268 using NIS+ & nscd 10.25s 7.83s 26.51s
4270 using NIS & old nscd [1] 13.83s 8.32s 29.60s
4272 Keep in mind that non-namelookup related operations dominate above times.
4273 It was just a common complain that using NIS+ unpacking the kernel is
4276 [1] The old nscd implementation is not available anymore since it was
4277 distributed with glibc up to version 2.0.98 and thus is now replaced.
4279 * Tim Waugh provided an implementation of the POSIX.2 wordexp function family.
4281 * Mark Kettenis provided a Hesiod NSS module.
4283 * The ELF dynamic loader knows how to process dynamic string tokens ($ORIGIN
4284 and $PLATFORM) in RPATHs and similar strings (Ulrich Drepper).
4286 * rcmd can now handle netgroups (Dick Streefland).
4288 * A port to the ARM was contributed by Phil Blundell, Pat Beirne and Scott
4291 * Support for the IPv6 protocol has been added to the socket API, as per the
4292 latest draft standards.
4294 * Support for Linux 2.2 has been added.
4296 * Interface changes relative to the latest 2.0.x release:
4297 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4298 addseverity NEW: Unix98
4299 alphasort64 NEW: LFS
4300 argp_err_exit_status NEW: argp, GNU ext
4301 argp_error NEW: argp, GNU ext
4302 argp_failure NEW: argp, GNU ext
4303 argp_help NEW: argp, GNU ext
4304 argp_parse NEW: argp, GNU ext
4305 argp_program_bug_address NEW: argp, GNU ext
4306 argp_program_version NEW: argp, GNU ext
4307 argp_program_version_hook NEW: argp, GNU ext
4308 argp_state_help NEW: argp, GNU ext
4309 argp_usage NEW: argp, GNU ext
4310 authdes_create NEW: Secure RPC
4311 authdes_getucred NEW: Secure RPC
4312 authdes_pk_create NEW: Secure RPC
4313 backtrace NEW: GNU ext.
4314 backtrace_symbols NEW: GNU ext.
4315 backtrace_symbols_fd NEW: GNU ext.
4317 cacosf NEW: ISO C 9x
4318 cacosh NEW: ISO C 9x
4319 cacoshf NEW: ISO C 9x
4320 cacoshl NEW: ISO C 9x
4321 cacosl NEW: ISO C 9x
4328 casinf NEW: ISO C 9x
4329 casinh NEW: ISO C 9x
4330 casinhf NEW: ISO C 9x
4331 casinhl NEW: ISO C 9x
4332 casinl NEW: ISO C 9x
4334 catanf NEW: ISO C 9x
4335 catanh NEW: ISO C 9x
4336 catanhf NEW: ISO C 9x
4337 catanhl NEW: ISO C 9x
4338 catanl NEW: ISO C 9x
4339 cbc_crypt NEW: Secure RPC
4343 ccoshf NEW: ISO C 9x
4344 ccoshl NEW: ISO C 9x
4350 cimagf NEW: ISO C 9x
4351 cimagl NEW: ISO C 9x
4352 clearerr_locked REMOVED
4353 clntunix_create NEW: sunrpc ext
4355 clog10 NEW: ISO C 9x
4356 clog10f NEW: ISO C 9x
4357 clog10l NEW: ISO C 9x
4367 cprojf NEW: ISO C 9x
4368 cprojl NEW: ISO C 9x
4370 crealf NEW: ISO C 9x
4371 creall NEW: ISO C 9x
4376 csinhf NEW: ISO C 9x
4377 csinhl NEW: ISO C 9x
4380 csqrtf NEW: ISO C 9x
4381 csqrtl NEW: ISO C 9x
4385 ctanhf NEW: ISO C 9x
4386 ctanhl NEW: ISO C 9x
4388 des_setparity NEW: Secure RPC
4389 ecb_crypt NEW: Secure RPC
4390 endutxent NEW: Unix98
4392 exp10f NEW: ISO C 9x
4393 exp10l NEW: ISO C 9x
4397 fattach NEW: STREAMS
4398 fdetach NEW: STREAMS
4402 feclearexcept NEW: ISO C 9x
4403 fegetenv NEW: ISO C 9x
4404 fegetexceptflag NEW: ISO C 9x
4405 fegetround NEW: ISO C 9x
4406 feholdexcept NEW: ISO C 9x
4408 feraiseexcept NEW: ISO C 9x
4409 ferror_locked REMOVED
4410 fesetenv NEW: ISO C 9x
4411 fesetexceptflag NEW: ISO C 9x
4412 fesetround NEW: ISO C 9x
4413 fetestexcept NEW: ISO C 9x
4414 feupdateenv NEW: ISO C 9x
4415 fflush_locked REMOVED
4419 fgets_unlocked NEW: GNU ext.
4420 fileno_locked REMOVED
4432 fputc_locked REMOVED
4433 fputs_unlocked NEW: GNU ext.
4434 fread_unlocked NEW: GNU ext.
4439 fstatvfs NEW: Unix98
4443 ftruncate64 NEW: LFS
4445 fwrite_unlocked NEW: GNU ext.
4446 gai_strerror NEW: IPv6
4450 getchar_locked REMOVED
4452 getdate_err NEW: Unix98
4453 getdate_r NEW: GNU ext.
4455 getnameinfo NEW: IPv6
4456 getnetname NEW: Secure RPC
4457 getpmsg NEW: STREAMS
4458 getpt NEW: Unix98 PTY
4459 getrlimit64 NEW: LFS
4460 getutxent NEW: Unix98
4461 getutxid NEW: Unix98
4462 getutxline NEW: Unix98
4463 glob64 NEW: GNU ext.
4464 globfree64 NEW: GNU ext.
4465 gnu_get_libc_release NEW: GNU ext.
4466 gnu_get_libc_version NEW: GNU ext.
4467 grantpt NEW: Unix98 PTY
4468 host2netname NEW: Secure RPC
4470 iconv_close NEW: iconv
4471 iconv_open NEW: iconv
4472 if_freenameindex NEW: IPv6
4473 if_indextoname NEW: IPv6
4474 if_nameindex NEW: IPv6
4475 if_nametoindex NEW: IPv6
4476 in6addr_any NEW: IPv6
4477 in6addr_loopback NEW: IPv6
4478 inet6_isipv4mapped NEW: IPv6
4479 isastream NEW: STREAMS
4480 iswblank NEW: GNU ext.
4481 key_decryptsession NEW: Secure RPC
4482 key_decryptsession_pk NEW: Secure RPC
4483 key_encryptsession NEW: Secure RPC
4484 key_encryptsession_pk NEW: Secure RPC
4485 key_gendes NEW: Secure RPC
4486 key_get_conv NEW: Secure RPC
4487 key_secretkey_is_set NEW: Secure RPC
4488 key_setnet NEW: Secure RPC
4489 key_setsecret NEW: Secure RPC
4490 llrint NEW: ISO C 9x
4491 llrintf NEW: ISO C 9x
4492 llrintl NEW: ISO C 9x
4493 llround NEW: ISO C 9x
4494 llroundf NEW: ISO C 9x
4495 llroundl NEW: ISO C 9x
4500 lrintf NEW: ISO C 9x
4501 lrintl NEW: ISO C 9x
4502 lround NEW: ISO C 9x
4503 lroundf NEW: ISO C 9x
4504 lroundl NEW: ISO C 9x
4506 makecontext NEW: Unix98
4507 mempcpy NEW: GNU ext.
4510 modify_ldt NEW: kernel
4514 nearbyint NEW: ISO C 9x
4515 nearbyintf NEW: ISO C 9x
4516 nearbyintl NEW: ISO C 9x
4517 netname2host NEW: Secure RPC
4518 netname2user NEW: Secure RPC
4519 nexttoward NEW: ISO C 9x
4520 nexttowardf NEW: ISO C 9x
4521 nexttowardl NEW: ISO C 9x
4525 passwd2des NEW: Secure RPC
4527 pow10f NEW: GNU ext.
4528 pow10l NEW: GNU ext.
4531 printf_size NEW: GNU ext.
4532 printf_size_info NEW: GNU ext.
4533 profil_counter REMOVED
4534 pthread_mutexattr_getkind_np REPLACED
4535 pthread_mutexattr_setkind_np REPLACED
4536 ptsname NEW: Unix98 PTY
4537 ptsname_r NEW: Unix98 PTY
4539 putchar_locked REMOVED
4540 putgrent NEW: GNU ext.
4542 putpmsg NEW: STREAMS
4543 pututxline NEW: Unix98
4547 readdir64_r NEW: LFS
4548 remquo NEW: ISO C 9x
4549 remquof NEW: ISO C 9x
4550 remquol NEW: ISO C 9x
4552 roundf NEW: ISO C 9x
4553 roundl NEW: ISO C 9x
4555 scalbln NEW: ISO C 9x
4556 scalblnf NEW: ISO C 9x
4557 scalblnl NEW: ISO C 9x
4559 sendfile NEW: kernel
4560 setcontext NEW: Unix98
4561 setrlimit64 NEW: LFS
4562 setutxent NEW: Unix98
4564 sigignore NEW: Unix98
4565 sigqueue NEW: POSIX.1b
4566 sigrelse NEW: Unix98
4567 sigset NEW: POSIX.1b
4568 sigtimedwait NEW: POSIX.1b
4569 sigwaitinfo NEW: POSIX.1b
4570 sincos NEW: GNU ext.
4571 sincosf NEW: GNU ext.
4572 sincosl NEW: GNU ext.
4576 strcasestr NEW: GNU ext.
4577 strtoimax NEW: ISO C 9x
4578 strtoumax NEW: ISO C 9x
4579 strverscmp NEW: GNU ext.
4580 svcauthdes_stats NEW: Secure RPC
4581 svcunix_create NEW: sunrpc etx
4582 svcunixfd_create NEW: sunrpc ext
4583 swapcontext NEW: Unix98
4584 tcgetsid NEW: Unix98 PTY
4585 tdestroy NEW: GNU ext.
4586 tgamma NEW: ISO C 9x
4587 tgammaf NEW: ISO C 9x
4588 tgammal NEW: ISO C 9x
4592 truncf NEW: ISO C 9x
4593 truncl NEW: ISO C 9x
4595 unlockpt NEW: Unix98 PTY
4596 updwtmpx NEW: Unix98
4597 user2netname NEW: Secure RPC
4598 utmpxname NEW: Unix98
4599 versionsort NEW: GNU ext.
4600 versionsort64 NEW: GNU ext.
4602 wcscasecmp NEW: GNU ext.
4603 wcsncasecmp NEW: GNU ext.
4604 wcsnlen NEW: GNU ext.
4605 wcstoimax NEW: ISO C 9x
4606 wcstoll NEW: ISO C 9x
4607 wcstoull NEW: ISO C 9x
4608 wcstoumax NEW: ISO C 9x
4610 wordexp NEW: POSIX.2
4611 wordfree NEW: POSIX.2
4612 write_profiling REMOVED
4613 xdecrypt NEW: Secure RPC
4614 xdr_authdes_cred NEW: Secure RPC
4615 xdr_authdes_verf NEW: Secure RPC
4616 xdr_cryptkeyarg NEW: Secure RPC
4617 xdr_cryptkeyarg2 NEW: Secure RPC
4618 xdr_cryptkeyres NEW: Secure RPC
4619 xdr_getcredres NEW: Secure RPC
4620 xdr_key_netstarg NEW: Secure RPC
4621 xdr_key_netstres NEW: Secure RPC
4622 xdr_keybuf NEW: Secure RPC
4623 xdr_keystatus NEW: Secure RPC
4624 xdr_netnamestr NEW: Secure RPC
4625 xdr_sizeof NEW: Secure RPC
4626 xdr_unixcred NEW: sunrpc ext
4627 xencrypt NEW: Secure RPC
4628 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4639 * inet_ntoa is thread-safe
4641 * updwtmp is moved from libutil to libc
4643 * rewrite of cbrt function
4645 * update of timezone data
4659 * add atoll function
4661 * fix complex problems in Berkeley DB code
4663 * fix math functions
4667 * fixed lots of header problems (especially Linux/GNU specific)
4669 * dynamic loader preserves all registers
4671 * Roland McGrath provided support for handling of auxiliary objects in
4672 the ELF dynamic loader.
4674 * support for parallel builds is improved
4678 * GNU extensions are no longer declared by default. To enable them you
4679 must define the macro `_GNU_SOURCE' in your program or compile with
4682 * The library has changed from using GNU ld symbol aliases to using weak
4683 symbols where available. The ELF object file format supports weak
4684 symbols; GNU ld also supports weak symbols in the a.out format. (There
4685 is also now support for other GNU ld extensions in ELF. Use the
4686 `--with-elf' option to configure to indicate you have ELF, and
4687 `--with-gnu-ld' if using GNU ld.) This change resulted in the deletion
4688 of many files which contained only symbol aliases, reducing the size of
4689 the source and the compiled library; many other files were renamed to
4690 less cryptic names previously occupied by the symbol alias files.
4691 There is a new header file <elf.h> for programs which operate on
4692 files in the ELF format.
4694 * Converted to Autoconf version 2, so `configure' has more options.
4695 Run `configure --help' to see the details.
4697 * The library can now be configured to build profiling, highly-optimized
4698 (but undebuggable), and/or shared libraries (ELF with GNU ld only). The
4699 `--enable-profile', `--enable-omitfp', and `--enable-shared' options to
4700 `configure' enable building these extra libraries. The shared library is
4701 built by default when using both ELF and GNU ld. When shared libraries
4702 are enabled, the new library `-ldl' is available for arbitrary run-time
4703 loading of shared objects; its interface is defined in <dlfcn.h>. The
4704 new header file <link.h> gives access to the internals of the run-time
4705 dynamic linker, `ld.so'. The shell script `ldd' is similar to the
4706 application of same name on other systems and it provides information
4707 about dynamically linked binaries.
4709 * The C library now provides the run-time support code for profiling
4710 executables compiled with `-pg'. Programs can control the profiling code
4711 through the interface in <sys/gmon.h>. The `gmon.out' files written by
4712 the GNU C library can be read only by GNU `gprof' (from GNU binutils);
4713 the support for this file format was contributed by David Mosberger-Tang.
4715 * The math code has been replaced with a math library based on fdlibm from
4716 Sun, and modified by JT Conklin and Ulrich Drepper with i387 support, by
4717 Ian Taylor with `float' functions and by Ulrich Drepper with `long double'
4718 functions. The math functions now reside in a separate library, so
4719 programs using them will need to use `-lm' their linking commands.
4721 * John C. Bowman contributed optimized ix87 assembler inline functions.
4723 * Ulrich Drepper has contributed support for an `/etc/nsswitch.conf'
4724 mechanism similar to that found in Solaris 2. This is now used for the
4725 group, passwd, hosts, networks, services, protocols, rpc, ethers,
4726 shadow, netgroup, publickey, and alias databases. The `nsswitch.conf'
4727 file controls what services are used for each individual database. This
4728 works by loading shared libraries with names specified in `nsswitch.conf',
4729 so service modules can be changed or added at any time without even
4730 relinking any program. Currently there are the file, db, and NIS based
4731 NSS services available.
4733 * The new functions `strtoq' and `strtouq' parse integer values from
4734 strings, like `strtol' and `strtoul', but they return `long long int' and
4735 `unsigned long long int' values, respectively (64-bit quantities).
4737 * The new functions `strtof' and `strtold' parse floating-point values from
4738 strings, like `strtod', but they return `float' and `long double' values,
4739 respectively (on some machines `double' and `long double' are the same).
4741 * Ulrich Drepper has contributed new implementations of the floating-point
4742 printing and reading code used in the `printf' family of functions and
4743 `strtod', `strtof', and `strtold'. These new functions are perfectly
4744 accurate, and much faster than the old ones.
4746 * The implementation of the POSIX locale model was completely rewritten by
4747 Ulrich Drepper. This includes the new programs `localedef' and `locale'
4748 to compile the POSIX locale definition.
4750 * The former dummy implementations of the strcoll and strxfrm function are
4751 now replaced by fully functional code contributed by Ulrich Drepper. The
4752 collation information comes from the POSIX locale definitions.
4754 * The new header <langinfo.h> defines an interface for accessing
4755 various locale-dependent data (using the locale chosen with `setlocale').
4757 * Ulrich Drepper has contributed a new suite of functions for operation on
4758 wide-character and multibyte-character strings, in <wchar.h>;
4759 and classification and case conversion of wide characters, in <wctype.h>.
4760 These new functions are conforming to the ISO C, Amendement 1 specification.
4762 * There is now a second implementation of the standard I/O library available.
4763 It comes from GNU libg++ as was written by Per Bothner, heavily modified
4764 by Hongjiu Lu and made thread safe by Ulrich Drepper.
4766 * You can now use positional parameter specifications in format strings
4767 for the `printf' and `scanf' families of functions. For example,
4768 `printf ("Number %2$d, Mr %1$s\n", "Jones", 6);'' prints
4769 ``Number 6, Mr Jones''. This is mainly useful when providing different
4770 format strings for different languages, whose grammars may dictate
4771 different orderings of the values being printed. To support this
4772 feature, the interface for `register_printf_handler' has changed; see
4773 the header file <printf.h> for details.
4775 * The `printf' and `scanf' families of functions now understand a new
4776 formatting flag for numeric conversions: the ' flag (e.g. %'d or %'f) says
4777 to group numbers as indicated by the locale; for `scanf' and friends, this
4778 says to accept as valid only a number with all the proper grouping
4779 separators in the right places. In the default "C" locale, numbers are
4780 not grouped; but locales for specific countries will define the usual
4781 conventions (i.e. separate thousands with `,' in the US locale).
4783 * The pgrp functions have been regularized, slightly incompatibly but much
4784 less confusingly. The core functions are now `getpgid' and `setpgid',
4785 which take arguments for the PID to operate on; the POSIX.1 `getpgrp' (no
4786 argument) and BSD `setpgrp' (identical to `setpgid') functions are
4787 provided for compatibility. There is no longer an incompatible `getpgrp'
4788 with an argument declared under _BSD_SOURCE; no BSD code uses it.
4790 * The new header file <fts.h> and suite of functions simplify programs that
4791 operate on directory trees. This code comes from 4.4 BSD.
4793 * The resolver code has been updated from the BIND 4.9.5-P1 release.
4794 Parts of the code were heavily modified by Ulrich Drepper to fit in the
4795 NSS scheme used in glibc.
4797 * There is a new malloc debugging hook `__memalign_hook'.
4799 * There are new typedefs `ushort' for `unsigned short int' and `uint' for
4800 `unsigned int' in <sys/types.h>. These are for compatibility only and
4801 their use is discouraged.
4803 * The `-lmcheck' library to enable standard malloc debugging hooks is now
4804 done differently, so that it works even without GNU ld.
4806 * New function `euidaccess' checks allowed access to a file like `access',
4807 but using the effective IDs instead of the real IDs.
4809 * The time zone data files have been updated for the latest and greatest
4810 local time conventions of the countries of the world.
4812 * The new function `dirfd' extracts the file descriptor used by a DIR stream;
4815 * The new functions `ecvt', `fcvt', and `gcvt' provide an obsolete interface
4816 for formatting floating-point numbers. They are provided only for
4817 compatibility; new programs should use `sprintf' instead. There are
4818 also equivalent function for the `long double' floating-point type and
4819 all functions also exist in a reentrant form.
4821 * The new auxiliary library `-lutil' from 4.4 BSD contains various
4822 functions for maintaining the login-record files (primarily of use to
4823 system programs such as `login'), and convenient functions for
4824 allocating and initializing a pseudo-terminal (pty) device.
4826 * Ulrich Drepper has contributed new support for System V style
4827 shared memory and IPC on systems that support it.
4829 * Ulrich Drepper has contributed several miscellaneous new functions found
4830 in System V: The `hsearch' family of functions provide an effective
4831 implementation of hash tables; `a64l' and `l64a' provide a very simple
4832 binary to ASCII mapping; `drand48' and friends provide a 48-bit random
4835 * Ulrich Drepper has contributed new reentrant counterparts for the
4836 `random' and `hsearch' families of functions; `random_r', `hsearch_r', etc.
4838 * Ulrich Drepper has contributed new, highly-optimized versions of several
4839 string functions for the i486/Pentium family of processors.
4841 * Ulrich Drepper has updated the Linux-specific code, based largely
4842 on work done in Hongjiu Lu's version of GNU libc for Linux.
4843 The GNU library now supports Linux versions 2.0.10 and later,
4844 using the ELF object file format (i[3456]86-*-linux).
4846 * Andreas Schwab has ported the C library to Linux/m68k (m68k-*-linux).
4848 * David Mosberger-Tang and Richard Henderson have ported the C library
4849 to Linux/Alpha (alpha-*-linux). Richard Henderson contributed the
4850 dynamic linking support for ELF/Alpha.
4852 * Richard Henderson contributed several Alpha optimized assembler function
4853 for arithmetic and string handling.
4855 * Ulrich Drepper has contributed a new set of message catalog functions to
4856 support multiple languages using the <libintl.h> interface, for use with
4857 his new package GNU gettext. Translation volunteers have contributed
4858 catalogs of the library's messages in Spanish, German, and Korean.
4860 * For compatibility with XPG4, Ulrich Drepper has contributed the `gencat'
4861 program and the `catgets' function for reading the catalog files it
4862 creates. (The <libintl.h> interface is preferred; we include the
4863 <nl_types.h> interface using `catgets' only for source compatibility with
4864 programs already written to use it.)
4866 * New header file <values.h> gives SVID-compatible names for <limits.h>
4869 * Various new macros, declarations, and small header files for compatibility
4872 * New function `group_member' is a convenient way to check if a process has
4873 a given effective group ID.
4875 * When using GCC 2.7 and later, the socket functions are now declared in a
4876 special way so that passing an argument of type `struct sockaddr_in *',
4877 `struct sockaddr_ns *', or `struct sockaddr_un *' instead of the generic
4878 `struct sockaddr *' type, does not generate a type-clash warning.
4880 * New function `error' declared in header file <error.h> is a convenient
4881 function for printing error messages and optionally exiting; this is the
4882 canonical function used in GNU programs. The new functions `err', `warn',
4883 and friends in header file <err.h> are the canonical 4.4 BSD interface for
4884 doing the same thing.
4886 * The <glob.h> interface has several new flags from 4.4 BSD that extend the
4887 POSIX.2 `glob' function to do ~ and {...} expansion.
4889 * New function `unsetenv' complements `setenv' for compatibility with 4.4 BSD.
4890 `clearenv' which is used in POSIX.9 is also available.
4892 * New function `getsid' returns session ID number on systems that support it.
4894 * We have incorporated the 4.4 BSD `db' library (version 1.85). New header
4895 files <db.h> and <mpool.h> provide a rich set of functions for several
4896 types of simple databases stored in memory and in files, and <ndbm.h> is
4897 an old `ndbm'-compatible interface using the `db' functions. Link with
4898 `-ldb' to get these functions.
4900 * New macro `strdupa' copies a string like `strdup', but uses local stack
4901 space from `alloca' instead of dynamic heap space from `malloc'.
4903 * New function `strnlen' is like `strlen' but searches only a given maximum
4904 number of characters for the null terminator. `stpncpy', `strndup' and
4905 `strndupa' are similar variants for the `stpcpy', `strdup' and `strdupa'
4908 * New function `statfs' in header <sys/statfs.h>.
4910 * The new <argz.h> and <envz.h> interfaces contributed by Miles Bader
4911 provide convenient functions for operating on blocks of null-terminated
4914 * A new suite of functions in <utmp.h> handle all the details of reading
4915 and writing the utmp file.
4917 * An implementation of the NIS/YP(tm) based NSS service was contributed by
4920 * Paul Eggert and Ulrich Drepper modified the `strftime' function to be
4921 completely POSIX compliant and also implemented the extended functionality
4922 to handle alternate digit representation and alternate era date formats.
4924 * Ulrich Drepper provided an implementation of the `strptime' function
4925 defined in XPG4.2 which transforms a string into a `struct tm' value.
4927 * Paul Eggert provided the tzselect shell script as part of the timezone
4928 code. The shell script makes it easy to select the correct timezone
4931 * The implementation of the malloc family of functions is completely replaced
4932 by a new implementation by Doug Lea with many improvements by Wolfram Gloger.
4933 The implementation uses the mmap function (if available) and it is
4934 optimized for the use in multi threaded programs.
4936 * Ulrich Drepper contributed a MD5 "encryption" for the crypt family of
4937 functions. This new functionality is usable by specifying a special
4938 salt string and it is compatible with implementation on *BSD systems.
4940 * Lots of functions from the XPG4.2 standard were added by Ulrich Drepper:
4941 `getsubopt' to handle second level command line options, `bsd_signal'
4942 to access BSD style `signal' functionality, the obsolete `regexp' style
4945 * the `lchown' function is available on system which support this
4948 * The implementation of the shadow password handling function was contributed
4951 * David Mosberger-Tang changed the SunRPC implementation to be 64bit safe.
4953 * POSIX.1g support was added. The <sys/select.h> header is available,
4954 `isfdtype' and `pselect' are implemented. Craig Metz contributed an
4955 implementation of `getaddrinfo'.
4959 * For cross-compilation you should now set `BUILD_CC' instead of `HOST_CC'.
4961 * New header file <fstab.h> and new functions `getfsspec', `getfsent' and
4962 friends, for parsing /etc/fstab. This code comes from 4.4 BSD.
4964 * The new function `daemon' from 4.4 BSD is useful for server programs that
4965 want to put themselves in the background.
4967 * Joel Sherrill has contributed support for several standalone boards that
4968 run without an operating system.
4970 * `printf', `scanf' and friends now accept a `q' type modifier for long
4971 long int as well as `ll'. Formats using these might be `%qu' or `%lld'.
4973 * All of the code taken from BSD (notably most of the math and networking
4974 routines) has been updated from the BSD 4.4-Lite release.
4976 * The resolver code has been updated from the BIND-4.9.3-BETA9 release.
4978 * The new functions `getdomainname' and `setdomainname' fetch or change the
4979 YP/NIS domain name. These are system calls which exist on systems which
4982 * The time zone data files have been updated for the latest international
4985 * The SunRPC programs `portmap' and `rpcinfo' are now installed in
4986 $(sbindir) (usually /usr/local/sbin) instead of $(bindir).
4990 * The C library now includes support for Sun RPC, from Sun's free
4991 RPCSRC-4.0 distribution. The `portmap', `rpcinfo', and `rpcgen' programs
4992 are included. (There is still no support for YP.)
4994 * Tom Quinn has contributed a port of the C library to SGI machines running
4995 Irix 4 (mips-sgi-irix4).
4997 * The new `lockf' function is a simplified interface to the locking
4998 facilities of `fcntl', included for compatibility.
5000 * New time functions `timegm', `timelocal', and `dysize' for compatibility.
5002 * New header file <sys/timeb.h> and new function `ftime' for compatibility.
5004 * New header files <poll.h> and <sys/poll.h> and new function `poll' for
5007 * The error message printed by `assert' for a failed assertion now includes
5008 the name of the program (if using GNU ld) and the name of the calling
5009 function (with versions of GCC that support this).
5011 * The `psignal' function is now declared in <signal.h>, not <stdio.h>.
5013 * The library now includes the <sys/mman.h> header file and memory
5014 management functions `mmap', `munmap', `mprotect', `msync', and
5015 `madvise', on systems that support those facilities.
5017 * The interface for `mcheck' has changed slightly: the function called to
5018 abort the program when an allocation inconsistency is detected now takes
5019 an argument that indicates the type of failure. The new function
5020 `mprobe' lets you request a consistency check for a particular block at
5021 any time (checks are normally done only when you call `free' or `realloc'
5024 * It is now possible to easily cross-compile the C library, building on one
5025 system a library to run on another machine and/or operating system. All
5026 you need to do is set the variable `HOST_CC' in `configparms' to the
5027 native compiler for programs to run on the machine you are building on (a
5028 few generator programs are used on Unix systems); set `CC' to the
5031 * The new function `fexecve' (only implemented on the GNU system) executes
5032 a program file given a file descriptor already open on the file.
5036 * Brendan Kehoe has contributed most of a port to the DEC Alpha
5037 running OSF/1 (alpha-dec-osf1). He says it is 75% complete.
5039 * You can set the variable `libprefix' in `configparms' to specify a prefix
5040 to be prepended to installed library files; this makes it easy to install
5041 the GNU C library to be linked as `-lgnuc' or whatever.
5043 * The new `stpncpy' is a cross between `stpcpy' and `strncpy': It
5044 copies a limited number of characters from a string, and returns the
5045 address of the last character written.
5047 * You no longer need to check for whether the installed `stddef.h' is
5048 compatible with the GNU C library. configure now checks for you.
5050 * You can now define a per-stream `fileno' function to convert the
5051 stream's cookie into an integral file descriptor.
5053 * ``malloc (0)'' no longer returns a null pointer. Instead, it
5054 allocates zero bytes of storage, and returns a unique pointer which
5055 you can pass to `realloc' or `free'. The behavior is undefined if
5056 you dereference this pointer.
5058 * The C library now runs on Sony NEWS m68k machines running either
5059 NewsOS 3 or NewsOS 4.
5061 * The new `syscall' function is a system-dependent primitive function
5062 for invoking system calls. It has the canonical behavior on Unix
5063 systems, including unreliable return values for some calls (such as
5064 `pipe', `fork' and `getppid').
5066 * The error code `EWOULDBLOCK' is now obsolete; it is always defined
5067 to `EAGAIN', which is the preferred name. On systems whose kernels
5068 use two distinct codes, the C library now translates EWOULDBLOCK to
5069 EAGAIN in every system call function.
5073 * The GNU C Library Reference Manual is now distributed with the library.
5074 `make dvi' will produce a DVI file of the printed manual.
5075 `make info' will produce Info files that you can read on line using C-h i
5076 in Emacs or the `info' program.
5077 Please send comments on the manual to bug-glibc-manual@gnu.org.
5079 * The library now supports SVR4 on i386s (i386-unknown-sysv4).
5081 * Brendan Kehoe has contributed a port to Sun SPARCs running Solaris 2.
5083 * Jason Merrill has contributed a port to the Sequent Symmetry running
5084 Dynix version 3 (i386-sequent-dynix).
5086 * The library has been ported to i386s running SCO 3.2.4 (also known as SCO
5087 ODT 2.0; i386-unknown-sco3.2.4) or SCO 3.2 (i386-unknown-sco3.2).
5089 * New function `memory_warnings' lets you arrange to get warnings when
5090 malloc is running out of memory to allocate, like Emacs gives you.
5092 * The C library now contains the relocating allocator used in Emacs 19 for
5093 its editing buffers. This allocator (ralloc) minimizes allocation
5094 overhead and fragmentation by moving allocated regions around whenever it
5095 needs to. You always refer to a ralloc'd region with a "handle" (a
5096 pointer to a pointer--an object of type `void **').
5098 * There is a new `printf' format: `%m' gives you the string corresponding
5099 to the error code in `errno'.
5101 * In `scanf' formats, you can now use `%as' or `%a[' to do the normal `%s'
5102 or `%[' conversion, but instead of filling in a fixed-sized buffer you
5103 pass, the `a' modifier says to fill in a `char **' you pass with a
5106 * The `fnmatch' function supports the new flag bits `FNM_LEADING_DIR' and
5107 `FNM_CASEFOLD'. `FNM_LEADING_DIR' lets a pattern like `foo*' match a
5108 name like `foo/bar'. `FNM_CASEFOLD' says to ignore case in matching.
5110 * `mkstemp' is a traditional Unix function to atomically create and open a
5111 uniquely-named temporary file.
5115 * The standard location for the file that says what the local timezone is
5116 has changed again. It is now `/usr/local/etc/localtime' (or more
5117 precisely, `${prefix}/etc/localtime') rather than `/etc/localtime'.
5119 * The distribution no longer contains any files with names longer than 14
5122 * `struct ttyent' has two new flag bits: TTY_TRUSTED and TTY_CONSOLE.
5123 These are set by the new `trusted' and `console' keywords in `/etc/ttys'.
5125 * New functions `ttyslot' and `syslog' from 4.4 BSD.
5129 * The configuration process has changed quite a bit. The `configure'
5130 script is now used just like the configuration scripts for other GNU
5131 packages. The `sysdeps' directory hierarchy is much rearranged.
5132 The file `INSTALL' explains the new scheme in detail.
5134 * The header files no longer need to be processed into ANSI C and
5135 traditional C versions. There is just one set of files to install, and
5136 it will work with ANSI or old C compilers (including `gcc -traditional').
5138 * Brendan Kehoe and Ian Lance Taylor have ported the library to the
5139 MIPS DECStation running Ultrix 4.
5141 * The Sun 4 startup code (crt0) can now properly load SunOS 4 shared libraries.
5142 Tom Quinn contributed the initial code. The GNU C library can NOT yet be
5143 made itself into a shared library.
5145 * Yet further improved support for the i386, running 4.3 BSD-like systems
5146 (such as Mach 3 with the Unix single-server), or System V.
5148 * New function `strncasecmp' to do case-insensitive string comparison
5149 with limited length.
5151 * New function `strsep' is a reentrant alternative to `strtok'.
5153 * New functions `scandir' and `alphasort' for searching directories.
5155 * New function `setenv' is a better interface to `putenv'.
5157 * Ian Lance Taylor has contributed an implementation of the SVID `ftw'
5158 function for traversing a directory tree.
5160 * The GNU obstack package is now also part of the C library.
5161 The new function `open_obstack_stream' creates a stdio stream that
5162 writes onto an obstack; `obstack_printf' and `obstack_vprintf' do
5163 formatted output directly to an obstack.
5165 * Miscellaneous new functions: reboot, nice, sigaltstack (4.4 BSD only),
5166 cfmakeraw, getusershell, getpass, swab, getttyent, seteuid, setegid.
5168 * `FNM_FILE_NAME' is another name for `FNM_PATHNAME', used with `fnmatch'.
5170 * The new functions `strfry' and `memfrob' do mysterious and wonderful
5171 things to your strings.
5173 * There are some new test programs: test-fseek, testmb, and testrand.
5175 * Some work has been done to begin porting the library to 4.4 BSD and Linux.
5176 These ports are not finished, but are a good starting place for really
5177 supporting those systems.
5179 * `/etc/localtime' is now the standard location for the file that says what
5180 the local timezone is, rather than `/usr/local/lib/zoneinfo/localtime'.
5181 This follows the general principle that `/etc' is the place for all local
5182 configuration files.
5184 * The C library header files now use `extern "C"' when used by the C++
5185 compiler, so the C library should now work with C++ code.
5187 * The header file <bstring.h> is gone. <string.h> now declares bcopy,
5188 bcmp, bzero, and ffs. (Update: nowadays these functions are declared
5191 * Mike Haertel (of GNU e?grep and malloc fame) has written a new sorting
5192 function which uses the `merge sort' algorithm, and is said to be
5193 significantly faster than the old GNU `qsort' function. Merge sort is
5194 now the standard `qsort' function. The new algorithm can require a lot
5195 of temporary storage; so, the old sorting function is called when the
5196 required storage is not available.
5198 * The C library now includes Michael Glad's Ultra Fast Crypt, which
5199 provides the Unix `crypt' function, plus some other entry points.
5201 * The code and header files taken from 4.4 BSD have been updated with the
5202 latest files released from Berkeley.
5204 ----------------------------------------------------------------------
5205 Copyright information:
5207 Copyright (C) 1992-2018 Free Software Foundation, Inc.
5209 Permission is granted to anyone to make or distribute verbatim copies
5210 of this document as received, in any medium, provided that the
5211 copyright notice and this permission notice are preserved,
5212 thus giving the recipient permission to redistribute in turn.
5214 Permission is granted to distribute modified versions
5215 of this document, or of portions of it,
5216 under the above conditions, provided also that they
5217 carry prominent notices stating who last changed them.
5220 version-control: never