]> git.ipfire.org Git - thirdparty/glibc.git/blob - NEWS
Revert "memusagestat: use local glibc when linking [BZ #18465]"
[thirdparty/glibc.git] / NEWS
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.
4
5 Please send GNU C library bug reports via <https://sourceware.org/bugzilla/>
6 using `glibc' in the "product" field.
7 \f
8 Version 2.28.1
9
10 Major new features:
11
12 * The entry for the new Japanese era has been added for ja_JP locale.
13
14 Deprecated and removed features, and other changes affecting compatibility:
15
16 * For powercp64le ABI, Transactional Lock Elision is now enabled iff kernel
17 indicates that it will abort the transaction prior to entering the kernel
18 (PPC_FEATURE2_HTM_NOSC on hwcap2). On older kernels the transaction is
19 suspended, and this caused some undefined side-effects issues by aborting
20 transactions manually. Glibc avoided it by abort transactions manually on
21 each syscall, but it lead to performance issues on newer kernels where the
22 HTM state is saved and restore lazily (the state being saved even when the
23 process actually does not use HTM).
24
25 The following bugs are resolved with this release:
26
27 [19444] build failures with -O1 due to -Wmaybe-uninitialized
28 [20018] getaddrinfo should reject IP addresses with trailing characters
29 [20209] localedata: Spelling mistake for Sunday in Greenlandic kl_GL
30 [22927] libanl: properly cleanup if first helper thread creation failed
31 [23400] stdlib/test-bz22786.c creates temporary files in glibc source tree
32 [23497] readdir64@GLIBC_2.1 cannot parse the kernel directory stream
33 [23509] CET enabled glibc is incompatible with the older linker
34 [23521] nss_files aliases database file stream leak
35 [23538] pthread_cond_broadcast: Fix waiters-after-spinning case
36 [23562] signal: Use correct type for si_band in siginfo_t
37 [23578] regex: Fix memory overread in re_compile_pattern
38 [23579] libc: Errors misreported in preadv2
39 [23606] Missing ENDBR32 in sysdeps/i386/start.S
40 [23614] powerpc: missing CFI register information in __mpn_* functions
41 [23679] gethostid: Missing NULL check for gethostbyname_r result
42 [23709] Fix CPU string flags for Haswell-type CPUs
43 [23717] Fix stack overflow in stdlib/tst-setcontext9
44 [23821] si_band in siginfo_t has wrong type long int on sparc64
45 [23822] ia64 static libm.a is missing exp2f, log2f and powf symbols
46 [23864] libc: [riscv] missing kernel-features.h undefines
47 [23844] pthread_rwlock_trywrlock results in hang
48 [23927] Linux if_nametoindex() does not close descriptor (CVE-2018-19591)
49 [23972] __old_getdents64 uses wrong d_off value on overflow
50 [24018] gettext may return NULL
51 [24022] riscv may lack <asm/syscalls.h>
52 [24024] strerror() might set errno to ENOMEM due to -fno-math-error
53 [24027] malloc: Integer overflow in realloc
54 [24034] tst-cancel21-static fails with SIGBUS on pre-ARMv7 when using GCC 8
55 [24040] riscv64: unterminated call chain in __thread_start
56 [24097] Can't use 64-bit register for size_t in assembly codes for x32 (CVE-2019-6488)
57 [24155] x32 memcmp can treat positive length as 0 (if sign bit in RDX is set) (CVE-2019-7309)
58 [24161] __run_fork_handlers self-deadlocks in malloc/tst-mallocfork2
59
60 Security related changes:
61
62 CVE-2018-19591: A file descriptor leak in if_nametoindex can lead to a
63 denial of service due to resource exhaustion when processing getaddrinfo
64 calls with crafted host names. Reported by Guido Vranken.
65
66 CVE-2019-6488: On x32, the size_t parameter may be passed in the lower
67 32 bits of a 64-bit register with with non-zero upper 32 bit. When it
68 happened, accessing the 32-bit size_t value as the full 64-bit register
69 in the assembly string/memory functions would cause a buffer overflow.
70 Reported by H.J. Lu.
71
72 CVE-2019-7309: x86-64 memcmp used signed Jcc instructions to check
73 size. For x86-64, memcmp on an object size larger than SSIZE_MAX
74 has undefined behavior. On x32, the size_t argument may be passed
75 in the lower 32 bits of the 64-bit RDX register with non-zero upper
76 32 bits. When it happened with the sign bit of RDX register set,
77 memcmp gave the wrong result since it treated the size argument as
78 zero. Reported by H.J. Lu.
79
80 CVE-2016-10739: The getaddrinfo function could successfully parse IPv4
81 addresses with arbitrary trailing characters, potentially leading to data
82 or command injection issues in applications.
83
84 CVE-2019-9169: Attempted case-insensitive regular-expression match
85 via proceed_next_node in posix/regexec.c leads to heap-based buffer
86 over-read. Reported by Hongxu Chen.
87 \f
88 Version 2.28
89
90 Major new features:
91
92 * The localization data for ISO 14651 is updated to match the 2016
93 Edition 4 release of the standard, this matches data provided by
94 Unicode 9.0.0. This update introduces significant improvements to the
95 collation of Unicode characters. This release deviates slightly from
96 the standard in that the collation element ordering for lowercase and
97 uppercase LATIN script characters is adjusted to ensure that regular
98 expressions with ranges like [a-z] and [A-Z] don't interleave e.g. A
99 is not matched by [a-z]. With the update many locales have been
100 updated to take advantage of the new collation information. The new
101 collation information has increased the size of the compiled locale
102 archive or binary locales.
103
104 * The GNU C Library can now be compiled with support for Intel CET, AKA
105 Intel Control-flow Enforcement Technology. When the library is built
106 with --enable-cet, the resulting glibc is protected with indirect
107 branch tracking (IBT) and shadow stack (SHSTK). CET-enabled glibc is
108 compatible with all existing executables and shared libraries. This
109 feature is currently supported on i386, x86_64 and x32 with GCC 8 and
110 binutils 2.29 or later. Note that CET-enabled glibc requires CPUs
111 capable of multi-byte NOPs, like x86-64 processors as well as Intel
112 Pentium Pro or newer. NOTE: --enable-cet has been tested for i686,
113 x86_64 and x32 on non-CET processors. --enable-cet has been tested
114 for x86_64 and x32 on CET SDVs, but Intel CET support hasn't been
115 validated for i686.
116
117 * The GNU C Library now has correct support for ABSOLUTE symbols
118 (SHN_ABS-relative symbols). Previously such ABSOLUTE symbols were
119 relocated incorrectly or in some cases discarded. The GNU linker can
120 make use of the newer semantics, but it must communicate it to the
121 dynamic loader by setting the ELF file's identification (EI_ABIVERSION
122 field) to indicate such support is required.
123
124 * Unicode 11.0.0 Support: Character encoding, character type info, and
125 transliteration tables are all updated to Unicode 11.0.0, using
126 generator scripts contributed by Mike FABIAN (Red Hat).
127
128 * <math.h> functions that round their results to a narrower type are added
129 from TS 18661-1:2014 and TS 18661-3:2015:
130
131 - fadd, faddl, daddl and corresponding fMaddfN, fMaddfNx, fMxaddfN and
132 fMxaddfNx functions.
133
134 - fsub, fsubl, dsubl and corresponding fMsubfN, fMsubfNx, fMxsubfN and
135 fMxsubfNx functions.
136
137 - fmul, fmull, dmull and corresponding fMmulfN, fMmulfNx, fMxmulfN and
138 fMxmulfNx functions.
139
140 - fdiv, fdivl, ddivl and corresponding fMdivfN, fMdivfNx, fMxdivfN and
141 fMxdivfNx functions.
142
143 * Two grammatical forms of month names are now supported for the following
144 languages: Armenian, Asturian, Catalan, Czech, Kashubian, Occitan, Ossetian,
145 Scottish Gaelic, Upper Sorbian, and Walloon. The following languages now
146 support two grammatical forms in abbreviated month names: Catalan, Greek,
147 and Kashubian.
148
149 * Newly added locales: Lower Sorbian (dsb_DE) and Yakut (sah_RU) also
150 include the support for two grammatical forms of month names.
151
152 * Building and running on GNU/Hurd systems now works without out-of-tree
153 patches.
154
155 * The renameat2 function has been added, a variant of the renameat function
156 which has a flags argument. If the flags are zero, the renameat2 function
157 acts like renameat. If the flag is not zero and there is no kernel
158 support for renameat2, the function will fail with an errno value of
159 EINVAL. This is different from the existing gnulib function renameatu,
160 which performs a plain rename operation in case of a RENAME_NOREPLACE
161 flags and a non-existing destination (and therefore has a race condition
162 that can clobber the destination inadvertently).
163
164 * The statx function has been added, a variant of the fstatat64
165 function with an additional flags argument. If there is no direct
166 kernel support for statx, glibc provides basic stat support based on
167 the fstatat64 function.
168
169 * IDN domain names in getaddrinfo and getnameinfo now use the system libidn2
170 library if installed. libidn2 version 2.0.5 or later is recommended. If
171 libidn2 is not available, internationalized domain names are not encoded
172 or decoded even if the AI_IDN or NI_IDN flags are passed to getaddrinfo or
173 getnameinfo. (getaddrinfo calls with non-ASCII names and AI_IDN will fail
174 with an encoding error.) Flags which used to change the IDN encoding and
175 decoding behavior (AI_IDN_ALLOW_UNASSIGNED, AI_IDN_USE_STD3_ASCII_RULES,
176 NI_IDN_ALLOW_UNASSIGNED, NI_IDN_USE_STD3_ASCII_RULES) have been
177 deprecated. They no longer have any effect.
178
179 * Parsing of dynamic string tokens in DT_RPATH, DT_RUNPATH, DT_NEEDED,
180 DT_AUXILIARY, and DT_FILTER has been expanded to support the full
181 range of ELF gABI expressions including such constructs as
182 '$ORIGIN$ORIGIN' (if valid). For SUID/GUID applications the rules
183 have been further restricted, and where in the past a dynamic string
184 token sequence may have been interpreted as a literal string it will
185 now cause a load failure. These load failures were always considered
186 unspecified behaviour from the perspective of the dynamic loader, and
187 for safety are now load errors e.g. /foo/${ORIGIN}.so in DT_NEEDED
188 results in a load failure now.
189
190 * Support for ISO C threads (ISO/IEC 9899:2011) has been added. The
191 implementation includes all the standard functions provided by
192 <threads.h>:
193
194 - thrd_current, thrd_equal, thrd_sleep, thrd_yield, thrd_create,
195 thrd_detach, thrd_exit, and thrd_join for thread management.
196
197 - mtx_init, mtx_lock, mtx_timedlock, mtx_trylock, mtx_unlock, and
198 mtx_destroy for mutual exclusion.
199
200 - call_once for function call synchronization.
201
202 - cnd_broadcast, cnd_destroy, cnd_init, cnd_signal, cnd_timedwait, and
203 cnd_wait for conditional variables.
204
205 - tss_create, tss_delete, tss_get, and tss_set for thread-local storage.
206
207 Application developers must link against libpthread to use ISO C threads.
208
209 Deprecated and removed features, and other changes affecting compatibility:
210
211 * The nonstandard header files <libio.h> and <_G_config.h> are no longer
212 installed. Software that was using either header should be updated to
213 use standard <stdio.h> interfaces instead.
214
215 * The stdio functions 'getc' and 'putc' are no longer defined as macros.
216 This was never required by the C standard, and the macros just expanded
217 to call alternative names for the same functions. If you hoped getc and
218 putc would provide performance improvements over fgetc and fputc, instead
219 investigate using (f)getc_unlocked and (f)putc_unlocked, and, if
220 necessary, flockfile and funlockfile.
221
222 * All stdio functions now treat end-of-file as a sticky condition. If you
223 read from a file until EOF, and then the file is enlarged by another
224 process, you must call clearerr or another function with the same effect
225 (e.g. fseek, rewind) before you can read the additional data. This
226 corrects a longstanding C99 conformance bug. It is most likely to affect
227 programs that use stdio to read interactive input from a terminal.
228 (Bug #1190.)
229
230 * The macros 'major', 'minor', and 'makedev' are now only available from
231 the header <sys/sysmacros.h>; not from <sys/types.h> or various other
232 headers that happen to include <sys/types.h>. These macros are rarely
233 used, not part of POSIX nor XSI, and their names frequently collide with
234 user code; see https://sourceware.org/bugzilla/show_bug.cgi?id=19239 for
235 further explanation.
236
237 <sys/sysmacros.h> is a GNU extension. Portable programs that require
238 these macros should first include <sys/types.h>, and then include
239 <sys/sysmacros.h> if __GNU_LIBRARY__ is defined.
240
241 * The tilegx*-*-linux-gnu configurations are no longer supported.
242
243 * The obsolete function ustat is no longer available to newly linked
244 binaries; the headers <ustat.h> and <sys/ustat.h> have been removed. This
245 function has been deprecated in favor of fstatfs and statfs.
246
247 * The obsolete function nfsservctl is no longer available to newly linked
248 binaries. This function was specific to systems using the Linux kernel
249 and could not usefully be used with the GNU C Library on systems with
250 version 3.1 or later of the Linux kernel.
251
252 * The obsolete function name llseek is no longer available to newly linked
253 binaries. This function was specific to systems using the Linux kernel
254 and was not declared in a header. Programs should use the lseek64 name
255 for this function instead.
256
257 * The AI_IDN_ALLOW_UNASSIGNED and NI_IDN_ALLOW_UNASSIGNED flags for the
258 getaddrinfo and getnameinfo functions have been deprecated. The behavior
259 previously selected by them is now always enabled.
260
261 * The AI_IDN_USE_STD3_ASCII_RULES and NI_IDN_USE_STD3_ASCII_RULES flags for
262 the getaddrinfo and getnameinfo functions have been deprecated. The STD3
263 restriction (rejecting '_' in host names, among other things) has been
264 removed, for increased compatibility with non-IDN name resolution.
265
266 * The fcntl function now have a Long File Support variant named fcntl64. It
267 is added to fix some Linux Open File Description (OFD) locks usage on non
268 LFS mode. As for others *64 functions, fcntl64 semantics are analogous with
269 fcntl and LFS support is handled transparently. Also for Linux, the OFD
270 locks act as a cancellation entrypoint.
271
272 * The obsolete functions encrypt, encrypt_r, setkey, setkey_r, cbc_crypt,
273 ecb_crypt, and des_setparity are no longer available to newly linked
274 binaries, and the headers <rpc/des_crypt.h> and <rpc/rpc_des.h> are no
275 longer installed. These functions encrypted and decrypted data with the
276 DES block cipher, which is no longer considered secure. Software that
277 still uses these functions should switch to a modern cryptography library,
278 such as libgcrypt.
279
280 * Reflecting the removal of the encrypt and setkey functions above, the
281 macro _XOPEN_CRYPT is no longer defined. As a consequence, the crypt
282 function is no longer declared unless _DEFAULT_SOURCE or _GNU_SOURCE is
283 enabled.
284
285 * The obsolete function fcrypt is no longer available to newly linked
286 binaries. It was just another name for the standard function crypt,
287 and it has not appeared in any header file in many years.
288
289 * We have tentative plans to hand off maintenance of the passphrase-hashing
290 library, libcrypt, to a separate development project that will, we hope,
291 keep up better with new passphrase-hashing algorithms. We will continue
292 to declare 'crypt' in <unistd.h>, and programs that use 'crypt' or
293 'crypt_r' should not need to change at all; however, distributions will
294 need to install <crypt.h> and libcrypt from a separate project.
295
296 In this release, if the configure option --disable-crypt is used, glibc
297 will not install <crypt.h> or libcrypt, making room for the separate
298 project's versions of these files. The plan is to make this the default
299 behavior in a future release.
300
301 Changes to build and runtime requirements:
302
303 GNU make 4.0 or later is now required to build glibc.
304
305 Security related changes:
306
307 CVE-2016-6261, CVE-2016-6263, CVE-2017-14062: Various vulnerabilities have
308 been fixed by removing the glibc-internal IDNA implementation and using
309 the system-provided libidn2 library instead. Originally reported by Hanno
310 Böck and Christian Weisgerber.
311
312 CVE-2017-18269: An SSE2-based memmove implementation for the i386
313 architecture could corrupt memory. Reported by Max Horn.
314
315 CVE-2018-11236: Very long pathname arguments to realpath function could
316 result in an integer overflow and buffer overflow. Reported by Alexey
317 Izbyshev.
318
319 CVE-2018-11237: The mempcpy implementation for the Intel Xeon Phi
320 architecture could write beyond the target buffer, resulting in a buffer
321 overflow. Reported by Andreas Schwab.
322
323 The following bugs are resolved with this release:
324
325 [1190] stdio: fgetc()/fread() behaviour is not POSIX compliant
326 [6889] manual: 'PWD' mentioned but not specified
327 [13575] libc: SSIZE_MAX defined as LONG_MAX is inconsistent with ssize_t,
328 when __WORDSIZE != 64
329 [13762] regex: re_search etc. should return -2 on memory exhaustion
330 [13888] build: /tmp usage during testing
331 [13932] math: dbl-64 pow unexpectedly slow for some inputs
332 [14092] nptl: Support C11 threads
333 [14095] localedata: Review / update collation data from Unicode / ISO
334 14651
335 [14508] libc: -Wformat warnings
336 [14553] libc: Namespace pollution loff_t in sys/types.h
337 [14890] libc: Make NT_PRFPREG canonical.
338 [15105] libc: Extra PLT references with -Os
339 [15512] libc: __bswap_constant_16 not compiled when -Werror -Wsign-
340 conversion is given
341 [16335] manual: Feature test macro documentation incomplete and out of
342 date
343 [16552] libc: Unify umount implementations in terms of umount2
344 [17082] libc: htons et al.: statement-expressions prevent use on global
345 scope with -O1 and higher
346 [17343] libc: Signed integer overflow in /stdlib/random_r.c
347 [17438] localedata: pt_BR: wrong d_fmt delimiter
348 [17662] libc: please implement binding for the new renameat2 syscall
349 [17721] libc: __restrict defined as /* Ignore */ even in c11
350 [17979] libc: inconsistency between uchar.h and stdint.h
351 [18018] dynamic-link: Additional $ORIGIN handling issues (CVE-2011-0536)
352 [18023] libc: extend_alloca is broken (questionable pointer comparison,
353 horrible machine code)
354 [18124] libc: hppa: setcontext erroneously returns -1 as exit code for
355 last constant.
356 [18471] libc: llseek should be a compat symbol
357 [18473] soft-fp: [powerpc-nofpu] __sqrtsf2, __sqrtdf2 should be compat
358 symbols
359 [18991] nss: nss_files skips large entry in database
360 [19239] libc: Including stdlib.h ends up with macros major and minor being
361 defined
362 [19463] libc: linknamespace failures when compiled with -Os
363 [19485] localedata: csb_PL: Update month translations + add yesstr/nostr
364 [19527] locale: Normalized charset name not recognized by setlocale
365 [19667] string: Missing Sanity Check for malloc calls in file 'testcopy.c'
366 [19668] libc: Missing Sanity Check for malloc() in file 'tst-setcontext-
367 fpscr.c'
368 [19728] network: out of bounds stack read in libidn function
369 idna_to_ascii_4i (CVE-2016-6261)
370 [19729] network: out of bounds heap read on invalid utf-8 inputs in
371 stringprep_utf8_nfkc_normalize (CVE-2016-6263)
372 [19818] dynamic-link: Absolute (SHN_ABS) symbols incorrectly relocated by
373 the base address
374 [20079] libc: Add SHT_X86_64_UNWIND to elf.h
375 [20251] libc: 32bit programs pass garbage in struct flock for OFD locks
376 [20419] dynamic-link: files with large allocated notes crash in
377 open_verify
378 [20530] libc: bswap_16 should use __builtin_bswap16() when available
379 [20890] dynamic-link: ldconfig: fsync the files before atomic rename
380 [20980] manual: CFLAGS environment variable replaces vital options
381 [21163] regex: Assertion failure in pop_fail_stack when executing a
382 malformed regexp (CVE-2015-8985)
383 [21234] manual: use of CFLAGS makes glibc detect no optimization
384 [21269] dynamic-link: i386 sigaction sa_restorer handling is wrong
385 [21313] build: Compile Error GCC 5.4.0 MIPS with -0S
386 [21314] build: Compile Error GCC 5.2.0 MIPS with -0s
387 [21508] locale: intl/tst-gettext failure with latest msgfmt
388 [21547] localedata: Tibetan script collation broken (Dzongkha and Tibetan)
389 [21812] network: getifaddrs() returns entries with ifa_name == NULL
390 [21895] libc: ppc64 setjmp/longjmp not fully interoperable with static
391 dlopen
392 [21942] dynamic-link: _dl_dst_substitute incorrectly handles $ORIGIN: with
393 AT_SECURE=1
394 [22241] localedata: New locale: Yakut (Sakha) locale for Russia (sah_RU)
395 [22247] network: Integer overflow in the decode_digit function in
396 puny_decode.c in libidn (CVE-2017-14062)
397 [22342] nscd: NSCD not properly caching netgroup
398 [22391] nptl: Signal function clear NPTL internal symbols inconsistently
399 [22550] localedata: es_ES locale (and other es_* locales): collation
400 should treat ñ as a primary different character, sync the collation
401 for Spanish with CLDR
402 [22638] dynamic-link: sparc: static binaries are broken if glibc is built
403 by gcc configured with --enable-default-pie
404 [22639] time: year 2039 bug for localtime etc. on 64-bit platforms
405 [22644] string: memmove-sse2-unaligned on 32bit x86 produces garbage when
406 crossing 2GB threshold (CVE-2017-18269)
407 [22646] localedata: redundant data (LC_TIME) for es_CL, es_CU, es_EC and
408 es_BO
409 [22735] time: Misleading typo in time.h source comment regarding
410 CLOCKS_PER_SECOND
411 [22753] libc: preadv2/pwritev2 fallback code should handle offset=-1
412 [22761] libc: No trailing `%n' conversion specifier in FMT passed from
413 `__assert_perror_fail ()' to `__assert_fail_base ()'
414 [22766] libc: all glibc internal dlopen should use RTLD_NOW for robust
415 dlopen failures
416 [22786] libc: Stack buffer overflow in realpath() if input size is close
417 to SSIZE_MAX (CVE-2018-11236)
418 [22787] dynamic-link: _dl_check_caller returns false when libc is linked
419 through an absolute DT_NEEDED path
420 [22792] build: tcb-offsets.h dependency dropped
421 [22797] libc: pkey_get() uses non-reserved name of argument
422 [22807] libc: PTRACE_* constants missing for powerpc
423 [22818] glob: posix/tst-glob_lstat_compat failure on alpha
424 [22827] dynamic-link: RISC-V ELF64 parser mis-reads flag in ldconfig
425 [22830] malloc: malloc_stats doesn't restore cancellation state on stderr
426 [22848] localedata: ca_ES: update date definitions from CLDR
427 [22862] build: _DEFAULT_SOURCE is defined even when _ISOC11_SOURCE is
428 [22884] math: RISCV fmax/fmin handle signalling NANs incorrectly
429 [22896] localedata: Update locale data for an_ES
430 [22902] math: float128 test failures with GCC 8
431 [22918] libc: multiple common of `__nss_shadow_database'
432 [22919] libc: sparc32: backtrace yields infinite backtrace with
433 makecontext
434 [22926] libc: FTBFS on powerpcspe
435 [22932] localedata: lt_LT: Update of abbreviated month names from CLDR
436 required
437 [22937] localedata: Greek (el_GR, el_CY) locales actually need ab_alt_mon
438 [22947] libc: FAIL: misc/tst-preadvwritev2
439 [22963] localedata: cs_CZ: Add alternative month names
440 [22987] math: [powerpc/sparc] fdim inlines errno, exceptions handling
441 [22996] localedata: change LC_PAPER to en_US in es_BO locale
442 [22998] dynamic-link: execstack tests are disabled when SELinux is
443 disabled
444 [23005] network: Crash in __res_context_send after memory allocation
445 failure
446 [23007] math: strtod cannot handle -nan
447 [23024] nss: getlogin_r is performing NSS lookups when loginid isn't set
448 [23036] regex: regex equivalence class regression
449 [23037] libc: initialize msg_flags to zero for sendmmsg() calls
450 [23069] libc: sigaction broken on riscv64-linux-gnu
451 [23094] localedata: hr_HR: wrong thousands_sep and mon_thousands_sep
452 [23102] dynamic-link: Incorrect parsing of multiple consecutive $variable
453 patterns in runpath entries (e.g. $ORIGIN$ORIGIN)
454 [23137] nptl: s390: pthread_join sometimes block indefinitely (on 31bit
455 and libc build with -Os)
456 [23140] localedata: More languages need two forms of month names
457 [23145] libc: _init/_fini aren't marked as hidden
458 [23152] localedata: gd_GB: Fix typo in "May" (abbreviated)
459 [23171] math: C++ iseqsig for long double converts arguments to double
460 [23178] nscd: sudo will fail when it is run in concurrent with commands
461 that changes /etc/passwd
462 [23196] string: __mempcpy_avx512_no_vzeroupper mishandles large copies
463 (CVE-2018-11237)
464 [23206] dynamic-link: static-pie + dlopen breaks debugger interaction
465 [23208] localedata: New locale - Lower Sorbian (dsb)
466 [23233] regex: Memory leak in build_charclass_op function in file
467 posix/regcomp.c
468 [23236] stdio: Harden function pointers in _IO_str_fields
469 [23250] nptl: Offset of __private_ss differs from GCC
470 [23253] math: tgamma test suite failures on i686 with -march=x86-64
471 -mtune=generic -mfpmath=sse
472 [23259] dynamic-link: Unsubstituted ${ORIGIN} remains in DT_NEEDED for
473 AT_SECURE
474 [23264] libc: posix_spawnp wrongly executes ENOEXEC in non compat mode
475 [23266] nis: stringop-truncation warning with new gcc8.1 in nisplus-
476 parser.c
477 [23272] math: fma(INFINITY,INFIITY,0.0) should be INFINITY
478 [23277] math: nan function should not have const attribute
479 [23279] math: scanf and strtod wrong for some hex floating-point
480 [23280] math: wscanf rounds wrong; wcstod is ok for negative numbers and
481 directed rounding
482 [23290] localedata: IBM273 is not equivalent to ISO-8859-1
483 [23303] build: undefined reference to symbol
484 '__parse_hwcap_and_convert_at_platform@@GLIBC_2.23'
485 [23307] dynamic-link: Absolute symbols whose value is zero ignored in
486 lookup
487 [23313] stdio: libio vtables validation and standard file object
488 interposition
489 [23329] libc: The __libc_freeres infrastructure is not properly run across
490 DSO boundaries.
491 [23349] libc: Various glibc headers no longer compatible with
492 <linux/time.h>
493 [23351] malloc: Remove unused code related to heap dumps and malloc
494 checking
495 [23363] stdio: stdio-common/tst-printf.c has non-free license
496 [23396] regex: Regex equivalence regression in single-byte locales
497 [23422] localedata: oc_FR: More updates of locale data
498 [23442] build: New warning with GCC 8
499 [23448] libc: Out of bounds access in IBM-1390 converter
500 [23456] libc: Wrong index_cpu_LZCNT
501 [23458] build: tst-get-cpu-features-static isn't added to tests
502 [23459] libc: COMMON_CPUID_INDEX_80000001 isn't populated for Intel
503 processors
504 [23467] dynamic-link: x86/CET: A property note parser bug
505 [24112] network: Do not send DNS queries for non-host names (where all
506 answers will be rejected)
507
508 \f
509 Version 2.27
510
511 Major new features:
512
513 * The GNU C Library can now be compiled with support for building static
514 PIE executables (See --enable-static-pie in INSTALL). These static PIE
515 executables are like static executables but can be loaded at any address
516 and provide additional security hardening benefits at the cost of some
517 memory and performance. When the library is built with --enable-static-pie
518 the resulting libc.a is usable with GCC 8 and above to create static PIE
519 executables using the GCC option '-static-pie'. This feature is currently
520 supported on i386, x86_64 and x32 with binutils 2.29 or later, and on
521 aarch64 with binutils 2.30 or later.
522
523 * Optimized x86-64 asin, atan2, exp, expf, log, pow, atan, sin, cosf,
524 sinf, sincosf and tan with FMA, contributed by Arjan van de Ven and
525 H.J. Lu from Intel.
526
527 * Optimized x86-64 trunc and truncf for processors with SSE4.1.
528
529 * Optimized generic expf, exp2f, logf, log2f, powf, sinf, cosf and sincosf.
530
531 * In order to support faster and safer process termination the malloc API
532 family of functions will no longer print a failure address and stack
533 backtrace after detecting heap corruption. The goal is to minimize the
534 amount of work done after corruption is detected and to avoid potential
535 security issues in continued process execution. Reducing shutdown time
536 leads to lower overall process restart latency, so there is benefit both
537 from a security and performance perspective.
538
539 * The abort function terminates the process immediately, without flushing
540 stdio streams. Previous glibc versions used to flush streams, resulting
541 in deadlocks and further data corruption. This change also affects
542 process aborts as the result of assertion failures.
543
544 * On platforms where long double has the IEEE binary128 format (aarch64,
545 alpha, mips64, riscv, s390 and sparc), the math library now implements
546 _Float128 interfaces for that type, as defined by ISO/IEC TS 18661-3:2015.
547 These are the same interfaces added in version 2.26 for some platforms where
548 this format is supported but is not the format of long double.
549
550 * On platforms with support for _Float64x (aarch64, alpha, i386, ia64,
551 mips64, powerpc64le, riscv, s390, sparc and x86_64), the math library now
552 implements interfaces for that type, as defined by ISO/IEC TS
553 18661-3:2015. These are corresponding interfaces to those supported for
554 _Float128.
555
556 * The math library now implements interfaces for the _Float32, _Float64 and
557 _Float32x types, as defined by ISO/IEC TS 18661-3:2015. These are
558 corresponding interfaces to those supported for _Float128.
559
560 * glibc now implements the memfd_create and mlock2 functions on Linux.
561
562 * Support for memory protection keys was added. The <sys/mman.h> header now
563 declares the functions pkey_alloc, pkey_free, pkey_mprotect, pkey_set,
564 pkey_get.
565
566 * The copy_file_range function was added.
567
568 * Optimized memcpy, mempcpy, memmove, and memset for sparc M7.
569
570 * The ldconfig utility now processes `include' directives using the C/POSIX
571 collation ordering. Previous glibc versions used locale-specific
572 ordering, the change might break systems that relied on that.
573
574 * Support for two grammatical forms of month names has been added.
575 In a call to strftime, the "%B" and "%b" format specifiers will now
576 produce the grammatical form required when the month is used as part
577 of a complete date. New "%OB" and "%Ob" specifiers produce the form
578 required when the month is named by itself. For instance, in Greek
579 and in many Slavic and Baltic languages, "%B" will produce the month
580 in genitive case, and "%OB" will produce the month in nominative case.
581
582 In a call to strptime, "%B", "%b", "%h", "%OB", "%Ob", and "%Oh"
583 are all valid and will all accept any known form of month
584 name---standalone or complete, abbreviated or full. In a call to
585 nl_langinfo, the query constants MON_1..12 and ABMON_1..12 return
586 the strings used by "%B" and "%b", respectively. New query
587 constants ALTMON_1..12 and _NL_ABALTMON_1..12 return the strings
588 used by "%OB" and "%Ob", respectively.
589
590 In a locale definition file, use "alt_mon" and "ab_alt_mon" to
591 define the strings for %OB and %Ob, respectively; these have the
592 same syntax as "mon" and "abmon". These arrays are optional; if they
593 are not provided then they have the same content as "mon" and "abmon",
594 respectively.
595
596 These features are provided for locales which define "alt_mon" and/or
597 "ab_alt_mon" in their locale source data. This release includes such
598 alternative month name data for the following languages: Belarusian,
599 Croatian, Greek, Lithuanian, Polish, Russian, and Ukrainian.
600
601 This feature is currently a GNU extension, but it is expected to
602 be added to the next revision of POSIX, and it is also already
603 available on some BSD-derived operating systems.
604
605 This feature will cause existing statically compiled applications
606 to fail to load locales and fall back to the builtin C/POSIX locales.
607 See notes below for other changes affecting compatibility.
608
609 * Support for the RISC-V ISA running on Linux has been added. This port
610 requires at least binutils-2.30, gcc-7.3.0, and linux-4.15; and is supported
611 for the following ISA and ABI pairs:
612
613 - rv64imac lp64
614 - rv64imafdc lp64
615 - rv64imafdc lp64d
616
617 Deprecated and removed features, and other changes affecting compatibility:
618
619 * Statically compiled applications attempting to load locales compiled for the
620 GNU C Library version 2.27 will fail and fall back to the builtin C/POSIX
621 locale. The reason for this is that the addition of the new "%OB" and "%Ob",
622 support for two grammatical forms of the month names, also extends the locale
623 data binary format. Static applications needing locale support must be
624 recompiled to match the runtime and data they are deployed with. In some
625 distributions there is an upgrade window where dynamically linked applications
626 may use a new library but the old locale data and also fall back to the
627 builtin C/POSIX locales; restarting the application process is sufficient to
628 fix this.
629
630 * Support for statically linked applications which call dlopen is deprecated
631 and will be removed in a future version of glibc. Applications which call
632 dlopen need to be linked dynamically instead.
633
634 * Support for old programs which use internal stdio data structures and
635 functions is deprecated. This includes programs which use the C++ streams
636 provided by libstdc++ in GCC 2.95. Programs which use the internal
637 symbols _IO_adjust_wcolumn, _IO_default_doallocate, _IO_default_finish,
638 _IO_default_pbackfail, _IO_default_uflow, _IO_default_xsgetn,
639 _IO_default_xsputn, _IO_doallocbuf, _IO_do_write, _IO_file_attach,
640 _IO_file_close, _IO_file_close_it, _IO_file_doallocate, _IO_file_fopen,
641 _IO_file_init, _IO_file_jumps, _IO_fileno, _IO_file_open,
642 _IO_file_overflow, _IO_file_read, _IO_file_seek, _IO_file_seekoff,
643 _IO_file_setbuf, _IO_file_stat, _IO_file_sync, _IO_file_underflow,
644 _IO_file_write, _IO_file_xsputn, _IO_flockfile, _IO_flush_all,
645 _IO_flush_all_linebuffered, _IO_free_backup_area, _IO_free_wbackup_area,
646 _IO_init, _IO_init_marker, _IO_init_wmarker, _IO_iter_begin, _IO_iter_end,
647 _IO_iter_file, _IO_iter_next, _IO_least_wmarker, _IO_link_in,
648 _IO_list_all, _IO_list_lock, _IO_list_resetlock, _IO_list_unlock,
649 _IO_marker_delta, _IO_marker_difference, _IO_remove_marker, _IO_seekmark,
650 _IO_seekwmark, _IO_str_init_readonly, _IO_str_init_static,
651 _IO_str_overflow, _IO_str_pbackfail, _IO_str_seekoff, _IO_str_underflow,
652 _IO_switch_to_main_wget_area, _IO_switch_to_wget_mode,
653 _IO_unsave_wmarkers, _IO_wdefault_doallocate, _IO_wdefault_finish,
654 _IO_wdefault_pbackfail, _IO_wdefault_setbuf, _IO_wdefault_uflow,
655 _IO_wdefault_xsgetn, _IO_wdefault_xsputn, _IO_wdoallocbuf, _IO_wdo_write,
656 _IO_wfile_jumps, _IO_wfile_overflow, _IO_wfile_sync, _IO_wfile_underflow,
657 _IO_wfile_xsputn, _IO_wmarker_delta, or _IO_wsetb may stop working with a
658 future version of glibc. Unlike other symbol removals, these old
659 applications will not be supported using compatibility symbols.
660
661 * On GNU/Linux, the obsolete Linux constant PTRACE_SEIZE_DEVEL is no longer
662 defined by <sys/ptrace.h>.
663
664 * libm no longer supports SVID error handling (calling a user-provided
665 matherr function on error) or the _LIB_VERSION variable to control error
666 handling. (SVID error handling and the _LIB_VERSION variable still work
667 for binaries linked against older versions of the GNU C Library.) The
668 libieee.a library is no longer provided. math.h no longer defines struct
669 exception, or the macros X_TLOSS, DOMAIN, SING, OVERFLOW, UNDERFLOW,
670 TLOSS, PLOSS and HUGE.
671
672 * The libm functions pow10, pow10f and pow10l are no longer supported for
673 new programs. Programs should use the standard names exp10, exp10f and
674 exp10l for these functions instead.
675
676 * The mcontext_t type is no longer the same as struct sigcontext. On
677 platforms where it was previously the same, this changes the C++ name
678 mangling for interfaces involving this type.
679
680 * The add-ons mechanism for building additional packages at the same time as
681 glibc has been removed. The --enable-add-ons configure option is now
682 ignored.
683
684 * The --without-fp configure option is now ignored. Whether hardware
685 floating-point instructions are used is now configured based on whether
686 the compiler used at configure time (without any options implied by a
687 --with-cpu= configure option) uses such instructions.
688
689 * The res_hnok, res_dnok, res_mailok and res_ownok functions now check that
690 the specified string can be parsed as a domain name.
691
692 * In the malloc_info output, the <heap> element may contain another <aspace>
693 element, "subheaps", which contains the number of sub-heaps.
694
695 * The libresolv function p_secstodate is no longer supported for new
696 programs.
697
698 * The tilepro-*-linux-gnu configuration is no longer supported.
699
700 * The nonstandard header files <libio.h> and <_G_config.h> are deprecated
701 and will be removed in a future release. Software that is still using
702 either header should be updated to use standard <stdio.h> interfaces
703 instead.
704
705 libio.h was originally the header for a set of supported GNU extensions,
706 but they have not been maintained as such in many years, they are now
707 standing in the way of improvements to stdio, and we don't think there are
708 any remaining external users. _G_config.h was never intended for public
709 use, but predates the bits convention.
710
711 Changes to build and runtime requirements:
712
713 * bison version 2.7 or later is required to generate code in the 'intl'
714 subdirectory.
715
716 Security related changes:
717
718 CVE-2009-5064: The ldd script would sometimes run the program under
719 examination directly, without preventing code execution through the
720 dynamic linker. (The glibc project disputes that this is a security
721 vulnerability; only trusted binaries must be examined using the ldd
722 script.)
723
724 CVE-2017-15670: The glob function, when invoked with GLOB_TILDE,
725 suffered from a one-byte overflow during ~ operator processing (either
726 on the stack or the heap, depending on the length of the user name).
727 Reported by Tim Rühsen.
728
729 CVE-2017-15671: The glob function, when invoked with GLOB_TILDE,
730 would sometimes fail to free memory allocated during ~ operator
731 processing, leading to a memory leak and, potentially, to a denial
732 of service.
733
734 CVE-2017-15804: The glob function, when invoked with GLOB_TILDE and
735 without GLOB_NOESCAPE, could write past the end of a buffer while
736 unescaping user names. Reported by Tim Rühsen.
737
738 CVE-2017-17426: The malloc function, when called with an object size near
739 the value SIZE_MAX, would return a pointer to a buffer which is too small,
740 instead of NULL. This was a regression introduced with the new malloc
741 thread cache in glibc 2.26. Reported by Iain Buclaw.
742
743 CVE-2017-1000408: Incorrect array size computation in _dl_init_paths leads
744 to the allocation of too much memory. (This is not a security bug per se,
745 it is mentioned here only because of the CVE assignment.) Reported by
746 Qualys.
747
748 CVE-2017-1000409: Buffer overflow in _dl_init_paths due to miscomputation
749 of the number of search path components. (This is not a security
750 vulnerability per se because no trust boundary is crossed if the fix for
751 CVE-2017-1000366 has been applied, but it is mentioned here only because
752 of the CVE assignment.) Reported by Qualys.
753
754 CVE-2017-16997: Incorrect handling of RPATH or RUNPATH containing $ORIGIN
755 for AT_SECURE or SUID binaries could be used to load libraries from the
756 current directory.
757
758 CVE-2018-1000001: Buffer underflow in realpath function when getcwd function
759 succeeds without returning an absolute path due to unexpected behaviour
760 of the Linux kernel getcwd syscall. Reported by halfdog.
761
762 CVE-2018-6485: The posix_memalign and memalign functions, when called with
763 an object size near the value of SIZE_MAX, would return a pointer to a
764 buffer which is too small, instead of NULL. Reported by Jakub Wilk.
765
766 CVE-2018-6551: The malloc function, when called with an object size near
767 the value of SIZE_MAX, would return a pointer to a buffer which is too
768 small, instead of NULL.
769
770 The following bugs are resolved with this release:
771
772 [866] glob: glob should match dangling symlinks
773 [1062] glob: sysdeps/generic/glob.c merge from gnulib (part 3 of 3)
774 [2522] localedata: ca_ES@valencia: new Valencian (meridional Catalan)
775 locale
776 [5997] math: Very slow execution of sinf function
777 [10580] localedata: hr_HR: updated locale
778 [10871] locale: 'mon' array should contain both nominative and genitive
779 cases
780 [12349] localedata: eu_ES: incorrect thousands separator
781 [13605] localedata: shn_MM: new Shan locale
782 [13805] localedata: ru_RU: currency should use ',' as radix point
783 [13953] localedata: km_KH: locale update
784 [13994] localedata: mjw_IN: new locale
785 [14121] build: make writes .mo files in po directory
786 [14333] libc: Fix the race between atexit() and exit()
787 [14681] dynamic-link: _dl_get_origin leaks memory via executable link map.
788 [14925] localedata: bn_*: LC_IDENTIFICATION.language key should be
789 "Bangla"
790 [15260] localedata: LC_MESSAGES.{yes,no}{str,expr}: various errors
791 [15261] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of full-
792 width Latin characters
793 [15332] localedata: es_CU: locale update
794 [15436] stdio: Don't close or flush stdio streams on abort
795 [15537] localedata: lv_LV: invalid collation for Latvian diacritical
796 letters
797 [16148] localedata: ca_ES: incorrect thousands separator
798 [16750] dynamic-link: ldd should not try to execute the binaries
799 (CVE-2009-5064)
800 [16777] localedata: pl_PL: incorrect thousands separator in locale
801 [16905] localedata: hanzi: new collation
802 [17563] localedata: cmn_TW: add hanzi collation
803 [17750] localedata: wrong collation order of diacritics in most locales
804 [17804] libc: scandirat fails with ENOMEM because it checks for errno even
805 if malloc succeeded
806 [17956] build: Build fails on missing definitions from header file
807 nss/nss.h when Mozilla NSS is used for cryptography
808 [18203] libc: realpath() does not handle unreachable paths correctly
809 [18572] dynamic-link: [arm] Lazy TLSDESC relocation has data race
810 [18812] localedata: kab_DZ: new Kabyle Algeria locale
811 [18822] libc: Internal functions are called via PLT
812 [18858] string: _HAVE_STRING_ARCH_xxx aren't defined for i386 nor x86_64
813 [19170] libc: __gmon_start__ defined in hppa in crtn.S
814 [19574] libc: glibc should support building static PIE binaries
815 [19852] localedata: charmaps/UTF-8: incorrect wcwidth for U+3099 and
816 U+309A
817 [19971] glob: glob: Do not skip entries with zero d_ino values
818 [19982] localedata: fr.po: spelling mistake for error code EXDEV
819 [20008] localedata: km_KH: convert to translit_neutral
820 [20009] localedata: tr_TR: convert LC_CTYPE to i18n
821 [20142] math: [x86_64] Add SSE4.1 trunc, truncf
822 [20204] dynamic-link: _dl_open_hook and _dlfcn_hook hardening
823 [20482] localedata: de_CH: abbreviated weekdays should be two letters
824 [20498] localedata: miq_NI: new Mískitu / Miskito (miq) language locale
825 for Nicaragua
826 [20532] nss: getaddrinfo uses errno and h_errno without guaranteeing
827 they're set, wrong errors returned by gaih_inet when lookup functions
828 are not found.
829 [20756] localedata: [PATCH] Use Unicode wise thousands separator
830 [20826] network: posix/tst-getaddrinfo5 fails on hosts without network
831 access
832 [20952] localedata: yuw_PG: new locale
833 [21084] localedata: charmaps/IBM858: new codepage
834 [21161] manual: [PATCH] fix typo in manual/arith.texi on strtoul prototype
835 [21242] libc: assert gives pedantic warning in old gcc versions
836 [21265] dynamic-link: _dl_runtime_resolve isn't compatible with Intel C++
837 __regcall calling convention
838 [21309] math: signed integer overflow in sysdeps/ieee754/dbl-64/e_pow.c
839 [21326] libc: C99 functions are not declared for C++11 and later if
840 _GNU_SOURCE is not predefined by g++
841 [21457] libc: sys/ucontext.h namespace
842 [21530] libc: tmpfile() should be implemented using O_TMPFILE
843 [21660] math: GCC fails to compile a formula with tgmath.h
844 [21672] nptl: sys-libs/glibc on ia64 crashes on thread exit: signal
845 SIGSEGV, Segmentation fault: pthread_create.c:432: __madvise
846 (pd->stackblock, freesize - PTHREAD_STACK_MIN, MADV_DONTNEED);
847 [21684] math: tgmath.h handling of complex integers
848 [21685] math: tgmath.h handling of bit-fields
849 [21686] math: tgmath.h handling of __int128
850 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
851 locale
852 [21745] libc: [powerpc64le] Extra PLT reference with --enable-stack-
853 protector=all
854 [21750] localedata: column width of characters incompatible with classical
855 wcwidth
856 [21754] malloc: malloc: Perform as little work as possible after heap
857 consistency check failures
858 [21780] libc: hppa: p{read,write}v2 does not set ENOSUP on invalid flag
859 [21790] libc: Missing __memset_zero_constant_len_parameter in libc.so
860 [21791] string: Unused XXX_chk_XXX functions in libc.a
861 [21815] dynamic-link: FAIL: elf/tst-prelink-cmp with GCC is defaulted to
862 PIE
863 [21836] localedata: Removed redundant data (LC_MONETARY) in various Indian
864 locales
865 [21845] localedata: Added new Locale bho_NP
866 [21853] localedata: Fix abday Which looks same as day in zh_SG
867 [21854] localedata: Added New Locale en_SC
868 [21864] libc: xmalloc.o is compiled with -DMODULE_NAME=libc
869 [21871] dynamic-link: _dl_runtime_resolve_avx_opt is slower than
870 _dl_runtime_resolve_avx_slow
871 [21885] network: getaddrinfo: gethosts does not release resolver context
872 on memory allocation failure
873 [21899] libc: XPG4.2 sigaction namespace
874 [21908] dynamic-link: dynamic linker broke on ia64 (mmap2 consolidation is
875 the suspect)
876 [21913] libc: static binaries SIGSEGV in __brk when host's gcc is pie-by-
877 default (i386)
878 [21915] nss: nss_files can return with NSS_STATUS_SUCCESS and a clobbered
879 errno value, causing getaddrinfo to fail
880 [21920] localedata: Fix p_cs_precedes/n_cs_precedes for mt_MT
881 [21922] network: getaddrinfo with AF_INET/AF_INET6 returns EAI_NONAME
882 instead of EAI_NODATA
883 [21928] libc: sys/ptrace.h: remove obsolete temporary development Linux
884 constant PTRACE_SEIZE_DEVEL
885 [21930] math: C-only gcc builtins used in <math.h> isinf
886 [21932] network: Unpaired __resolv_context_get in generic get*_r
887 implementation
888 [21941] math: powerpc: Wrong register constraint for xssqrtqp in sqrtf128
889 [21944] libc: sigval namespace
890 [21951] localedata: Update hanzi collation by stroke
891 [21955] math: Wrong alignment of L(SP_RANGE)/L(SP_INF_0) in
892 sysdeps/x86_64/fpu/e_expf.S
893 [21956] libc: Stack allocation in MIPS syscall impl (ubounded stack
894 allocation in syscall loops)
895 [21959] localedata: Fix Country name for xh_ZA
896 [21960] localedata: Fix abmon for bem_ZM
897 [21966] math: AVX2 mathvec functions use FMA without checking
898 [21967] math: When 512-bit AVX2 wrapper functions in mathvec are used?
899 [21971] localedata: Added New Locale for mfe_MU
900 [21972] libc: assert macro requires operator== (int) for its argument type
901 [21973] math: [sparc] libm missing sqrtl compat symbol
902 [21974] libc: Remove __bb_init_func and __bb_exit_func
903 [21982] string: stratcliff.c: error: assuming signed overflow does not
904 occur with -O3
905 [21986] stdio: __guess_grouping is called incorrectly
906 [21987] math: [sparc32] wrong bits/long-double.h installed
907 [22019] localedata: Wrong placement of monetary symbol in el_GR (negative
908 amounts)
909 [22022] localedata: Missing country_name for mni_IN
910 [22023] localedata: Removed redundant data (LC_TIME and LC_MESSAGES) for
911 niu_NZ
912 [22025] locale: iconv: Inconsistency between pointer mangling and NULL
913 checks
914 [22026] locale: iconv_open: heap overflow on gconv_init failure
915 [22028] math: bits/math-finite.h _MSUF_ expansion namespace
916 [22035] math: [m68k] bits/math-inline.h macro namespace
917 [22038] localedata: Fix abbreviated weeks and months for Somali
918 [22044] localedata: Remove redundant data for Limburgish Language
919 [22050] malloc: Linking with -lmcheck does not hook
920 __malloc_initialize_hook correctly
921 [22051] libc: zero terminator in the middle of glibc's .eh_frame
922 [22052] malloc: malloc failed to compile with GCC 7 and -O3
923 [22070] localedata: charmaps/UTF-8: wcwidth for
924 Prepended_Concatenation_Mark codepoints set to 0 (should be 1)
925 [22074] localedata: charmaps/UTF-8: wcwidth for U+1160-U+11FF (Hangul
926 Jungseong and Jongseong) should be 0
927 [22078] nss: nss_files performance issue in multi mode
928 [22082] math: bits/math-finite.h exp10 condition
929 [22086] libc: pcprofiledump incorrect cross-endian condition
930 [22093] dynamic-link: ld.so no longer searches in .../x86_64
931 [22095] network: Name server address allocation memory leak in resolv.conf
932 parsing after OOM
933 [22096] network: __resolv_conf_attach can incorrectly free passed conf
934 object
935 [22100] localedata: om_KE: LC_TIME: copy redundant data from om_ET
936 [22101] dynamic-link: Dynamic loader must ignore "debug" shared objects
937 e.g. ET_GNU_DEBUG_*
938 [22111] malloc: malloc: per thread cache is not returned when thread exits
939 [22112] localedata: Fix LC_TELEPHONE/LC_NAME for az_AZ
940 [22134] libc: [linux] implement fexecve with execveat
941 [22142] libc: [powerpc] printf oupts a wrong value of DBL_MAX on ppc64 and
942 ppc64le
943 [22145] libc: ttyname() gives up too early in the face of namespaces
944 [22146] math: C++ build issue with float128 on x86_64
945 [22153] nptl: nptl: save error code before process termination
946 [22156] libc: [hppa,ia64,microblaze] Executable stack default
947 [22159] malloc: malloc: MALLOC_CHECK_ broken with --enable-tunables=no
948 [22161] nscd: nscd cache prune for netgroups hangs after timeout bump
949 [22165] libc: [hppa] Text relocations in libc.so
950 [22180] libc: destructor registered via __cxa_atexit is called twice
951 [22183] glob: commit 5554304f0ddd ("posix: Allow glob to match dangling
952 symlinks") cause "make" segfaults
953 [22189] math: [powerpc] math_private.h definitions of math_opt_barrier and
954 math_force_eval
955 [22207] libc: FAIL: stdlib/test-atexit-race
956 [22225] math: nearbyint arithmetic moved before feholdexcept
957 [22229] math: [sparc32] missing copysignl, fabsl, fmal compat symbols
958 [22235] math: iscanonical in C++ and float128
959 [22243] math: log2(0) and log10(0) are wrong in downward rounding without
960 the svid compat wrapper
961 [22244] math: ynf and yn are wrong without the svid compat wrapper
962 [22273] libc: Improper assert in Linux posix_spawn implementation
963 [22284] libc: -pg -pie doesn't work
964 [22292] locale: localedef exits with error 4 when it should be error 1
965 [22294] locale: Allow "" for int_currency_symbol definition in locales.
966 [22295] locale: Don't warn on non-symbolic characters in locale sources in
967 --verbose.
968 [22296] math: glibc 2.26: signbit build issue with Gcc 5.5.0 on x86_64
969 [22298] nptl: x32: lockups on recursive pthread_mutex_lock after upgrade
970 to 2.26
971 [22299] dynamic-link: Problem with $PLATFORM on x86_64 platform
972 [22320] glob: Buffer overflow in glob with GLOB_TILDE (CVE-2017-15670)
973 [22321] libc: sysconf(_SC_IOV_MAX) returns -1 on Linux
974 [22322] libc: [mips64] wrong bits/long-double.h installed
975 [22325] glob: Memory leak in glob with GLOB_TILDE (CVE-2017-15671)
976 [22332] glob: Buffer overflow in glob with GLOB_TILDE in unescaping
977 (CVE-2017-15804)
978 [22336] localedata: cs_CZ LC_COLLATE does not use i18n
979 [22343] malloc: Integer overflow in posix_memalign
980 [22347] libc: getrandom() returns the number of bytes that were copied to
981 the buffer even though the comments say "Return 0 on success and -1 on
982 failure."
983 [22353] string: sysdeps/i386/i586/strcpy.S isn't maintainable
984 [22362] libc: Installed crt1.o, crti,.o and crtn.o files are used with
985 -m32
986 [22370] dynamic-link: Incorrect note padding check
987 [22375] libc: malloc returns pointer from tcache_get when should return
988 NULL (CVE-2017-17426)
989 [22377] math: iseqsig, float128 and C++
990 [22382] localedata: Error in tpi_PG locale
991 [22387] localedata: Replace unicode sequences <Uxxxx> for characters
992 inside the ASCII printable range
993 [22402] math: [powerpc64le] __MATH_TG does not support _Float128 for
994 -mlong-double-64
995 [22403] localedata: Slash needs escaping in some locales
996 [22408] malloc: malloc_info access heaps without arena lock, ignores heaps
997 [22409] network: res_hnok does not accept some host names used on the
998 Internet
999 [22412] network: res_dnok, res_hnok should perform syntax checks
1000 [22413] network: ns_name_pton ignores syntactically invalid trailing
1001 backslash
1002 [22415] stdio: setvbuf can lead to invalid free/segfault
1003 [22432] build: Non-deterministic build
1004 [22439] malloc: malloc_info should compute summary statistics for all sub-
1005 heaps in an arena
1006 [22442] network: if_nametoindex could report index for the wrong
1007 networking interface
1008 [22446] build: aliasing violation calling readlink in handle_request
1009 [22447] build: unsafe call to strlen with a non-string in getlogin_r.c
1010 [22457] libc: Generic preadv/pwritev incorrectly calls __posix_memalign
1011 [22459] libc: FAIL: elf/check-localplt with __stack_chk_fail related to
1012 __nscd_hash/__nss_hash
1013 [22463] network: p_secstodate overflow handling
1014 [22469] localedata: pl_PL LC_COLLATE does not use i18n
1015 [22478] libc: sigwait can fail with EINTR
1016 [22505] libc: ldconfig processes include directive in locale-specific
1017 order
1018 [22515] localedata: hsb_DE LC_COLLATE does not use copy "iso14651_t1"
1019 [22517] localedata: et_EE LC_COLLATE does not use copy "iso14651_t1"
1020 [22519] localedata: is_IS LC_COLLATE does not use copy "iso14651_t1"
1021 [22524] localedata: lt_LT LC_COLLATE does not use copy "iso14651_t1"
1022 [22527] localedata: tr_TR LC_COLLATE does not use copy "iso14651_t1"
1023 [22534] localedata: Collation rules for Serbian and Bosnian should be the
1024 same as for Croatian
1025 [22561] math: [DR#471] cacosh (0 + iNaN) should return NaN +/- i pi/2
1026 [22568] math: [DR#471] ctanh (0 + iNaN), ctanh (0 + i Inf)
1027 [22577] libc: missing newline after "cannot allocate TLS data structures
1028 for initial thread"
1029 [22588] manual: manual/conf.texi: missing underscore in front of
1030 SC_SSIZE_MAX
1031 [22593] math: nextafter and nexttoward are declared with const attribute
1032 [22596] manual: manual: finite(nan) wrongly described as returning nonzero
1033 [22603] string: ia64 memchr overflows internal pointer check
1034 [22605] libc: SH clone does not set the exit code correctly
1035 [22606] dynamic-link: Incorrect array size computation in _dl_init_paths
1036 (CVE-2017-1000408)
1037 [22607] dynamic-link: Buffer Overflow in _dl_init_paths (CVE-2017-1000409)
1038 [22611] malloc: malloc/tst-realloc wrongly assumes that errno must not be
1039 modified in case of success
1040 [22614] build: gcc: error: unrecognized command line option ‘-no-pie’
1041 [22615] manual: manual: ambiguous wording about errno value in case of
1042 success
1043 [22624] libc: MIPS setjmp() saves incorrect 'o0' register in --enable-
1044 stack-protector=all
1045 [22625] dynamic-link: RPATH $ORIGIN replaced by PWD for AT_SECURE/SUID
1046 binaries or if /proc is not mounted (CVE-2017-16997)
1047 [22627] dynamic-link: $ORIGIN in $LD_LIBRARY_PATH is substituted twice
1048 [22630] build: $(no-pie-ldflag) is no longer effective
1049 [22631] math: [m68k] Bad const attributes in bits/mathinline.h
1050 [22635] nptl: pthread_self returns NULL before libpthread is loaded
1051 [22636] nptl: PTHREAD_STACK_MIN is too small on x86-64
1052 [22637] nptl: guard size is subtracted from thread stack size instead of
1053 adding it on top
1054 [22648] libc: getrlimit/setrlimit with RLIM_INFINITY broken on alpha
1055 [22657] localedata: hu_HU: Avoid double space in date
1056 [22660] math: fmax, fmin sNaN handling on alpha
1057 [22664] libc: New warning of GCC8
1058 [22665] math: alpha: ceil and floor raise inexact exceptions
1059 [22666] math: alpha: trunc raise inexact exceptions
1060 [22667] libc: makecontext lacks stack alignment on i386
1061 [22678] libc: prlimit fails for RLIM_INFINITY values on 32-bit machines
1062 [22679] libc: getcwd(3) can succeed without returning an absolute path
1063 (CVE-2018-1000001)
1064 [22685] libc: PowerPC: Static AT_SECURE binaries segfault with lock-
1065 elision and tunables
1066 [22687] math: [powerpc-nofpu] complex long double functions spurious
1067 "invalid" exception
1068 [22688] math: [powerpc-nofpu] remainderl wrong sign of zero result
1069 [22690] math: [ldbl-128ibm] lrintl, lroundl missing "invalid" exceptions
1070 [22691] math: [powerpc-nofpu] fmaxmagl, fminmagl spurious "invalid"
1071 exception
1072 [22693] math: [ldbl-128ibm] log1pl (-qNaN) spurious "invalid" exception
1073 [22697] math: [powerpc] llround spurious "inexact" exceptions on 32-bit
1074 power4
1075 [22701] nis: Incomplete removal of libnsl
1076 [22702] math: [powerpc-nofpu] nearbyintl traps with trapping "inexact"
1077 [22707] libc: Missing defines in elf.h for DF_1_STUB and DF_1_PIE.
1078 [22715] dynamic-link: FAIL: elf/tst-audit10
1079 [22719] libc: Backtrace tests fail on hppa
1080 [22742] libc: [aarch64] mcontext_t __reserved field got renamed
1081 [22743] nptl: __pthread_register_cancel corrupts stack after f81ddabffd
1082 [22765] crypt: (struct crypt_data *data)->initialized is not set to zero
1083 before the first call to crypt_r () in crypt/badsalttest.c
1084
1085 \f
1086 Version 2.26
1087
1088 Major new features:
1089
1090 * A per-thread cache has been added to malloc. Access to the cache requires
1091 no locks and therefore significantly accelerates the fast path to allocate
1092 and free small amounts of memory. Refilling an empty cache requires locking
1093 the underlying arena. Performance measurements show significant gains in a
1094 wide variety of user workloads. Workloads were captured using a special
1095 instrumented malloc and analyzed with a malloc simulator. Contributed by
1096 DJ Delorie with the help of Florian Weimer, and Carlos O'Donell.
1097
1098 * Unicode 10.0.0 Support: Character encoding, character type info, and
1099 transliteration tables are all updated to Unicode 10.0.0, using
1100 generator scripts contributed by Mike FABIAN (Red Hat).
1101 These updates cause user visible changes, especially the changes in
1102 wcwidth for many emoji characters cause problems when emoji sequences
1103 are rendered with pango, see for example:
1104 https://bugzilla.gnome.org/show_bug.cgi?id=780669#c5
1105
1106 * Collation of Hungarian has been overhauled and is now consistent with "The
1107 Rules of Hungarian Orthography, 12th edition" (Bug 18934). Contributed by
1108 Egmont Koblinger.
1109
1110 * Improvements to the DNS stub resolver, contributed by Florian Weimer:
1111
1112 - The GNU C Library will now detect when /etc/resolv.conf has been
1113 modified and reload the changed configuration. The new resolver option
1114 “no-reload” (RES_NORELOAD) disables this behavior.
1115
1116 - The GNU C Library now supports an arbitrary number of search domains
1117 (configured using the “search” directive in /etc/resolv.conf);
1118 previously, there was a hard limit of six domains. For backward
1119 compatibility, applications that directly modify the ‘_res’ global
1120 object are still limited to six search domains.
1121
1122 - When the “rotate” (RES_ROTATE) resolver option is active, the GNU C
1123 Library will now randomly pick a name server from the configuration as a
1124 starting point. (Previously, the second name server was always used.)
1125
1126 * The tunables feature is now enabled by default. This allows users to tweak
1127 behavior of the GNU C Library using the GLIBC_TUNABLES environment variable.
1128
1129 * New function reallocarray, which resizes an allocated block (like realloc)
1130 to the product of two sizes, with a guaranteed clean failure upon integer
1131 overflow in the multiplication. Originally from OpenBSD, contributed by
1132 Dennis Wölfing and Rüdiger Sonderfeld.
1133
1134 * New wrappers for the Linux-specific system calls preadv2 and pwritev2.
1135 These are extended versions of preadv and pwritev, respectively, taking an
1136 additional flags argument. The set of supported flags depends on the
1137 running kernel; full support currently requires kernel 4.7 or later.
1138
1139 * posix_spawnattr_setflags now supports the flag POSIX_SPAWN_SETSID, to
1140 create a new session ID for the spawned process. This feature is
1141 scheduled to be added to the next major revision of POSIX; for the time
1142 being, it is available under _GNU_SOURCE.
1143
1144 * errno.h is now safe to use from C-preprocessed assembly language on all
1145 supported operating systems. In this context, it will only define the
1146 Exxxx constants, as preprocessor macros expanding to integer literals.
1147
1148 * On ia64, powerpc64le, x86-32, and x86-64, the math library now implements
1149 128-bit floating point as defined by ISO/IEC/IEEE 60559:2011 (IEEE
1150 754-2008) and ISO/IEC TS 18661-3:2015. Contributed by Paul E. Murphy,
1151 Gabriel F. T. Gomes, Tulio Magno Quites Machado Filho, and Joseph Myers.
1152
1153 To compile programs that use this feature, the compiler must support
1154 128-bit floating point with the type name _Float128 (as defined by TS
1155 18661-3) or __float128 (the nonstandard name used by GCC for C++, and for
1156 C prior to version 7). _GNU_SOURCE or __STDC_WANT_IEC_60559_TYPES_EXT__
1157 must be defined to make the new interfaces visible.
1158
1159 The new functions and macros correspond to those present for other
1160 floating-point types (except for a few obsolescent interfaces not
1161 supported for the new type), with F128 or f128 suffixes; for example,
1162 strtof128, HUGE_VAL_F128 and cosf128. Following TS 18661-3, there are no
1163 printf or scanf formats for the new type; the strfromf128 and strtof128
1164 interfaces should be used instead.
1165
1166 Deprecated and removed features, and other changes affecting compatibility:
1167
1168 * The synchronization that pthread_spin_unlock performs has been changed to
1169 now be equivalent to a C11 atomic store with release memory order to the
1170 spin lock's memory location. Previously, several (but not all)
1171 architectures used stronger synchronization (e.g., containing what is
1172 often called a full barrier). This change can improve performance, but
1173 may affect odd fringe uses of spin locks that depend on the previous
1174 behavior (e.g., using spin locks as atomic variables to try to implement
1175 Dekker's mutual exclusion algorithm).
1176
1177 * The port to Native Client running on ARMv7-A (--host=arm-nacl) has been
1178 removed.
1179
1180 * Sun RPC is deprecated. The rpcgen program, librpcsvc, and Sun RPC headers
1181 will only be built and installed when the GNU C Library is configured with
1182 --enable-obsolete-rpc. This allows alternative RPC implementations, such
1183 as TIRPC or rpcsvc-proto, to be used.
1184
1185 * The NIS(+) name service modules, libnss_nis, libnss_nisplus, and
1186 libnss_compat, are deprecated, and will not be built or installed by
1187 default.
1188
1189 The NIS(+) support library, libnsl, is also deprecated. By default, a
1190 compatibility shared library will be built and installed, but not headers
1191 or development libraries. Only a few NIS-related programs require this
1192 library. (In particular, the GNU C Library has never required programs
1193 that use 'gethostbyname' to be linked with libnsl.)
1194
1195 Replacement implementations based on TIRPC, which additionally support
1196 IPv6, are available from <https://github.com/thkukuk/>. The configure
1197 option --enable-obsolete-nsl will cause libnsl's headers, and the NIS(+)
1198 name service modules, to be built and installed.
1199
1200 * The DNS stub resolver no longer performs EDNS fallback. If EDNS or DNSSEC
1201 support is enabled, the configured recursive resolver must support EDNS.
1202 (Responding to EDNS-enabled queries with responses which are not
1203 EDNS-enabled is fine, but FORMERR responses are not.)
1204
1205 * res_mkquery and res_nmkquery no longer support the IQUERY opcode. DNS
1206 servers have not supported this opcode for a long time.
1207
1208 * The _res_opcodes variable has been removed from libresolv. It had been
1209 exported by accident.
1210
1211 * <string.h> no longer includes inline versions of any string functions,
1212 as this kind of optimization is better done by the compiler. The macros
1213 __USE_STRING_INLINES and __NO_STRING_INLINES no longer have any effect.
1214
1215 * The nonstandard header <xlocale.h> has been removed. Most programs should
1216 use <locale.h> instead. If you have a specific need for the definition of
1217 locale_t with no other declarations, please contact
1218 libc-alpha@sourceware.org and explain.
1219
1220 * The obsolete header <sys/ultrasound.h> has been removed.
1221
1222 * The obsolete signal constant SIGUNUSED is no longer defined by <signal.h>.
1223
1224 * The obsolete function cfree has been removed. Applications should use
1225 free instead.
1226
1227 * The stack_t type no longer has the name struct sigaltstack. This changes
1228 the C++ name mangling for interfaces involving this type.
1229
1230 * The ucontext_t type no longer has the name struct ucontext. This changes
1231 the C++ name mangling for interfaces involving this type.
1232
1233 * On M68k GNU/Linux and MIPS GNU/Linux, the fpregset_t type no longer has
1234 the name struct fpregset. On Nios II GNU/Linux, the mcontext_t type no
1235 longer has the name struct mcontext. On SPARC GNU/Linux, the struct
1236 mc_fq, struct rwindow, struct fpq and struct fq types are no longer
1237 defined in sys/ucontext.h, the mc_fpu_t type no longer has the name struct
1238 mc_fpu, the gwindows_t type no longer has the name struct gwindows and the
1239 fpregset_t type no longer has the name struct fpu. This changes the C++
1240 name mangling for interfaces involving those types.
1241
1242 * On S/390 GNU/Linux, the constants defined by <sys/ptrace.h> have been
1243 synced with the kernel:
1244
1245 - PTRACE_GETREGS, PTRACE_SETREGS, PTRACE_GETFPREGS and PTRACE_SETFPREGS
1246 are not supported on this architecture and have been removed.
1247
1248 - PTRACE_SINGLEBLOCK, PTRACE_SECCOMP_GET_FILTER, PTRACE_PEEKUSR_AREA,
1249 PTRACE_POKEUSR_AREA, PTRACE_GET_LAST_BREAK, PTRACE_ENABLE_TE,
1250 PTRACE_DISABLE_TE and PTRACE_TE_ABORT_RAND have been added.
1251
1252 Programs that assume the GET/SETREGS ptrace requests are universally
1253 available will now fail to build, instead of malfunctioning at runtime.
1254
1255 Changes to build and runtime requirements:
1256
1257 * Linux kernel 3.2 or later is required at runtime, on all architectures
1258 supported by that kernel. (This is a change from version 2.25 only for
1259 x86-32 and x86-64.)
1260
1261 * GNU Binutils 2.25 or later is now required to build the GNU C Library.
1262
1263 * On most architectures, GCC 4.9 or later is required to build the GNU C
1264 Library. On powerpc64le, GCC 6.2 or later is required.
1265
1266 Older GCC versions and non-GNU compilers are still supported when
1267 compiling programs that use the GNU C Library. (We do not know exactly
1268 how old, and some GNU extensions to C may be _de facto_ required. If you
1269 are interested in helping us make this statement less vague, please
1270 contact libc-alpha@sourceware.org.)
1271
1272 Security related changes:
1273
1274 * The DNS stub resolver limits the advertised UDP buffer size to 1200 bytes,
1275 to avoid fragmentation-based spoofing attacks (CVE-2017-12132).
1276
1277 * LD_LIBRARY_PATH is now ignored in binaries running in privileged AT_SECURE
1278 mode to guard against local privilege escalation attacks (CVE-2017-1000366).
1279
1280 * Avoid printing a backtrace from the __stack_chk_fail function since it is
1281 called on a corrupt stack and a backtrace is unreliable on a corrupt stack
1282 (CVE-2010-3192).
1283
1284 * A use-after-free vulnerability in clntudp_call in the Sun RPC system has been
1285 fixed (CVE-2017-12133).
1286
1287 The following bugs are resolved with this release:
1288
1289 [984] network: Respond to changed resolv.conf in gethostbyname
1290 [5010] network: sunrpc service cleanup causes unwanted port mapper traffic
1291 [12068] localedata: sc_IT: misspelled yesexpr/day/abday/mon/abmon/date_fmt
1292 fields
1293 [12189] libc: __stack_chk_fail should not attempt a backtrace
1294 (CVE-2010-3192)
1295 [14096] time: Race condition on timezone/tst-timezone.out
1296 [14172] localedata: az_IR: new locale
1297 [14995] build: glibc fails to build if gold is the default linker, even if
1298 ld.bfd is available
1299 [15998] build: [powerpc] Set arch_minimum_kernel for powerpc LE
1300 [16637] network: inet_pton function is accepting IPv6 with bad format
1301 [16640] string: string/strtok.c: undefined behaviour inconsistent between
1302 x86 and other generic code
1303 [16875] localedata: ko_KR: fix lang_name
1304 [17225] localedata: ar_SY: localized month names for May and June are
1305 incorrect
1306 [17297] localedata: da_DK: wrong date_fmt string
1307 [18907] stdio: Incorrect order of __wur __THROW in <printf.h>
1308 [18934] localedata: hu_HU: collate: fix multiple bugs and add tests
1309 [18988] nptl: pthread wastes memory with mlockall(MCL_FUTURE)
1310 [19066] localedata: ar_SA abbreviated day and month names are in English
1311 [19569] network: resolv: Support an arbitrary number of search domains
1312 [19570] network: Implement random DNS server selection in the stub
1313 resolver
1314 [19838] locale: localedef fails on PA-RISC
1315 [19919] localedata: iso14651_t1_common: Correct the Malayalam sorting
1316 order of 0D36 and 0D37
1317 [19922] localedata: iso14651_t1_common: Define collation for Malayalam
1318 chillu characters
1319 [20098] libc: FAIL: debug/backtrace-tst on hppa
1320 [20257] network: sunrpc: clntudp_call does not enforce timeout when
1321 receiving data
1322 [20275] localedata: locale day/abday/mon/abmon should not have trailing
1323 whitespace
1324 [20313] localedata: Update locale data to Unicode 9.0
1325 [20424] manual: Document how to provide a malloc replacement
1326 [20496] localedata: agr_PE: new language locale Awajún / Aguaruna (agr)
1327 for Peru
1328 [20686] locale: Add el_GR@euro to SUPPORTED.
1329 [20831] dynamic-link: _dl_map_segments does not test for __mprotect
1330 failures consistently
1331 [21015] dynamic-link: Document and fix --enable-bind-now
1332 [21016] nptl: pthread_cond support is broken on hppa
1333 [21029] libc: glibc-2.23 (and later) fails to compile with -fno-omit-
1334 frame-pointer on i386
1335 [21049] libc: segfault in longjmp_chk() due to clobbered processor
1336 register
1337 [21075] libc: unused assigment to %g4 in sparc/sparc{64,32}/clone.S
1338 [21088] libc: Build fails with --enable-static-nss
1339 [21094] math: cosf(1.57079697) has 3 ulp error on targets where the
1340 generic c code is used
1341 [21109] libc: Tunables broken on big-endian
1342 [21112] math: powf has large ulp errors with base close to 1 and exponent
1343 around 4000
1344 [21115] network: sunrpc: Use-after-free in error path in clntudp_call
1345 (CVE-2017-12133)
1346 [21120] malloc: glibc malloc is incompatible with GCC 7
1347 [21130] math: Incorrect return from y0l (-inf) and y1l (-inf) when linking
1348 with -lieee
1349 [21134] math: Exception (divide by zero) not set for y0/y1 (0.0) and y0/y1
1350 (-0.0) when linking with -lieee
1351 [21171] math: log10, log2 and lgamma return incorrect results
1352 [21179] libc: handle R_SPARC_DISP64 and R_SPARC_REGISTER relocs
1353 [21182] libc: __memchr_sse2: regression in glibc-2.25 on i686
1354 [21207] localedata: ce_RU: update weekdays from CLDR
1355 [21209] dynamic-link: LD_HWCAP_MASK read in setuid binaries
1356 [21217] localedata: Update months from CLDR-31
1357 [21232] libc: miss posix_fadvise64 on MIPS64 when static linking
1358 [21243] libc: support_delete_temp_file should issue warning for failed
1359 remove()
1360 [21244] libc: support resolv_test_start() socket fd close should be
1361 checked for errors.
1362 [21253] libc: localedef randomly segfaults when using -fstack-check due to
1363 new posix_spawn implementation
1364 [21258] dynamic-link: Branch predication in _dl_runtime_resolve_avx512_opt
1365 leads to lower CPU frequency
1366 [21259] libc: [alpha] termios.h missing IXANY for POSIX
1367 [21261] libc: [sparc64] bits/setjmp.h namespace
1368 [21267] network: [mips] bits/socket.h IOC* namespace
1369 [21268] libc: [alpha] termios.h NL2, NL3 namespace
1370 [21270] libc: mmap64 silently truncates large offset values
1371 [21275] libc: posix_spawn always crashes on ia64 now
1372 [21277] libc: [alpha] termios.h missing IUCLC for UNIX98 and older
1373 [21280] math: [powerpc] logbl for POWER7 return incorrect results
1374 [21289] libc: Incorrect declaration for 32-bit platforms with
1375 _FILE_OFFSET_BITS=64 causes build error
1376 [21295] network: GETAI(AF_UNSPEC) drops IPv6 addresses if nss module does
1377 not support gethostbyname4_r
1378 [21298] nptl: rwlock can deadlock on frequent reader/writer phase
1379 switching
1380 [21338] malloc: mallopt M_ARENA_MAX doesn't set the maximum number of
1381 arenas
1382 [21340] libc: Support POSIX_SPAWN_SETSID
1383 [21357] libc: unwind-dw2-fde deadlock when using AddressSanitizer
1384 [21359] network: ns_name_pack needs additional byte in destination buffer
1385 [21361] network: resolv: Reduce advertised EDNS0 buffer size to guard
1386 against fragmentation attacks (CVE-2017-12132)
1387 [21369] network: resolv: Remove EDNS fallback
1388 [21371] libc: Missing timespec definition when compiled with _XOPEN_SOURCE
1389 and _POSIX_C_SOURCE
1390 [21386] nptl: Assertion in fork for distinct parent PID is incorrect
1391 [21391] dynamic-link: x86: Set dl_platform and dl_hwcap from CPU features
1392 [21393] stdio: Missing dup3 error check in freopen, freopen64
1393 [21396] libc: Use AVX2 memcpy/memset on Skylake server
1394 [21399] localedata: Bad description for U00EC in
1395 localedata/charmaps/CP1254
1396 [21411] malloc: realloc documentation error
1397 [21426] network: sys/socket.h uio.h namespace
1398 [21428] libc: [aarch64] tst-backtrace5 testsuite failure
1399 [21445] libc: signal.h bsd_signal namespace
1400 [21455] network: Network headers stdint.h namespace
1401 [21474] network: resolv: res_init does not use RES_DFLRETRY (2) but 4 for
1402 retry value
1403 [21475] network: resolv: Overlong search path is truncated mid-label
1404 [21511] libc: sigstack namespace
1405 [21512] libc: clone() ends up calling exit_group() through _exit() wrapper
1406 [21514] libc: sysdeps/unix/sysv/linux/sys/syscall.h:31:27: fatal error:
1407 bits/syscall.h: No such file or directory
1408 [21517] libc: struct sigaltstack namespace
1409 [21528] dynamic-link: Duplicated minimal strtoul implementations in ld.so
1410 [21533] localedata: Update locale data to Unicode 10.0
1411 [21537] libc:
1412 ../sysdeps/unix/sysv/linux/s390/s390-32/__makecontext_ret.S:44: Error:
1413 junk at end of line, first unrecognized character is `@'
1414 [21538] libc: SIG_HOLD missing for XPG4
1415 [21539] libc: S390: Mismatch between kernel and glibc ptrace.h with
1416 request 12: PTRACE_SINGLEBLOCK vs PTRACE_GETREGS.
1417 [21542] libc: Use conservative default for sysconf (_SC_NPROCESSORS_ONLN)
1418 [21543] libc: sigevent namespace
1419 [21548] libc: [mips] get/set/make/swap context for MIPS O32 assume wrong
1420 size for general purpose registers in mcontext_t structure
1421 [21550] libc: sigwait namespace
1422 [21552] libc: XPG4 bsd_signal namespace
1423 [21554] libc: sigpause namespace
1424 [21560] libc: sys/wait.h signal.h namespace
1425 [21561] libc: waitid namespace
1426 [21573] nptl: GCC 7: /usr/bin/install: cannot remove
1427 '/usr/include/stdlib.h': Permission denied
1428 [21575] libc: sys/wait.h missing struct rusage definition
1429 [21584] libc: sigaltstack etc namespace
1430 [21597] libc: siginterrupt namespace
1431 [21607] math: hppa: FAIL: math/test-tgmath
1432 [21609] dynamic-link: Incomplete workaround for GCC __tls_get_addr ABI
1433 issue on x86-64
1434 [21622] libc: [tile] missing SA_* for POSIX.1:2008
1435 [21624] dynamic-link: ld.so: Unsafe alloca allows local attackers to alias
1436 stack and heap (CVE-2017-1000366)
1437 [21625] libc: wait3 namespace
1438 [21654] nss: Incorrect pointer alignment in NSS group merge result
1439 construction
1440 [21657] network: Parse interface zone id for node-local multicast
1441 [21662] string: memcmp-avx2-movbe.S lacks saturating subtraction for
1442 between_2_3
1443 [21666] libc: .symver is used on common symbol
1444 [21668] network: resolv: res_init cross-thread broadcast introduces race
1445 conditions
1446 [21687] math: tgmath.h totalorder, totalordermag return type
1447 [21694] locale: Current Glibc Locale Does Not Support Tok-Pisin and Fiji
1448 Hindi Locale
1449 [21696] libc: Incorrect assumption of of __cpu_mask in
1450 posix/sched_cpucount.c
1451 [21697] libc: sysdeps/posix/spawni.c: 2 * suspicious condition ?
1452 [21706] localedata: yesstr and nostr are missing for Breton [LC_MESSAGES]
1453 locale
1454 [21707] math: ppc64le: Invalid IFUNC resolver from libgcc calls getauxval,
1455 leading to relocation crash
1456 [21709] libc: resolv_conf.c:552: update_from_conf: Assertion
1457 `resolv_conf_matches (resp, conf)' failed.
1458 [21710] localedata: Added Samoan language locale for Samoa
1459 [21711] localedata: Pashto yesstr/nostr locale are missing
1460 [21715] nptl: sysdeps/nptl/bits/pthreadtypes.h: typedef guard
1461 __have_pthread_attr_t can cause redefinition of typedef ‘pthread_attr_t’
1462 [21721] localedata: Incorrect Full Weekday names for ks_IN@devanagari
1463 [21723] localedata: yesstr/nostr missing for Chinese language locale
1464 [21724] localedata: yesstr and nostr are missing for Xhosa [LC_MESSAGES]
1465 locale
1466 [21727] localedata: yesstr and nostr are missing for Tsonga [LC_MESSAGES]
1467 locale
1468 [21728] localedata: New Locale for Tongan language
1469 [21729] localedata: incorrect LC_NAME fields for hi_IN
1470 [21733] localedata: yesstr and nostr are missing for zh_HK
1471 [21734] localedata: Missing yesstr and nostr are for kw_GB
1472 [21738] libc: misc/tst-preadvwritev2 and misc/tst-preadvwritev64v2 fail
1473 [21741] libc: Undefined __memmove_chk_XXX and __memset_chk_XXX in libc.a
1474 [21742] libc: _dl_num_cache_relocations is undefined in libc.a
1475 [21743] localedata: ks_IN@devanagari: abday strings mismatch the day
1476 strings
1477 [21744] libc: Tests failing on --enable-tunables --enable-stack-
1478 protector=all
1479 [21749] localedata: Wrong abbreviated day name (“abday”) for
1480 ar_JO/ar_LB/ar_SY
1481 [21756] localedata: missing yesstr, nostr for nds_DE and nds_NL
1482 [21757] localedata: missing yesstr, nostr for pap_AW and pap_CW
1483 [21759] localedata: missing yesstr and nostr for Tigrinya
1484 [21760] localedata: Fix LC_MESSAGES and LC_ADDRESS for anp_IN
1485 [21766] localedata: Wrong LC_MESSAGES for om_ET Locale
1486 [21767] localedata: Missing Bislama locales
1487 [21768] localedata: Missing yesstr and nostr for aa_ET
1488 [21770] localedata: Missing Field in li_NL
1489 [21778] nptl: Robust mutex may deadlock
1490 [21779] libc: MicroBlaze segfaults when loading libpthread
1491 [21783] localedata: Fix int_select international_call_prefixes
1492 [21784] localedata: Inconsistency in country_isbn
1493 [21788] localedata: Missing Country Postal Abbreviations
1494 [21794] localedata: Added-country_isbn-for-Italy
1495 [21795] localedata: Add/Fix country_isbn for France
1496 [21796] localedata: Added country_isbn for Republic of Korea
1497 [21797] localedata: Fix inconsistency in country_isbn and missing prefixes
1498 [21799] localedata: Added int_select international_call_prefixes
1499 [21801] localedata: Added int_select international_call_prefixes
1500 [21804] nptl: Double semicolon in thread-shared-types.h
1501 [21807] localedata: LC_ADDRESS fix for pap_CW
1502 [21808] localedata: Fix LC_ADDRESS for pap_AW
1503 [21821] localedata: Added country_name in mai_IN
1504 [21822] localedata: Fix LC_TIME for mai_IN
1505 [21823] localedata: missing yesstr, nostr for sa_IN
1506 [21825] localedata: Fix name_mrs for mag_IN
1507 [21828] localedata: 2.26 changelog should mention user visible changes
1508 with unicode 9.0
1509 [21835] localedata: Added Maithili language locale for Nepal
1510 [21838] localedata: Removed redundant data for the_NP
1511 [21839] localedata: Fix LC_MONETARY for ta_LK
1512 [21844] localedata: Fix Latin characters and Months Sequence.
1513 [21848] localedata: Fix mai_NP Title Name
1514
1515 \f
1516 Version 2.25
1517
1518 * The feature test macro __STDC_WANT_LIB_EXT2__, from ISO/IEC TR
1519 24731-2:2010, is supported to enable declarations of functions from that
1520 TR. Note that not all functions from that TR are supported by the GNU C
1521 Library.
1522
1523 * The feature test macro __STDC_WANT_IEC_60559_BFP_EXT__, from ISO/IEC TS
1524 18661-1:2014, is supported to enable declarations of functions and macros
1525 from that TS. Note that not all features from that TS are supported by
1526 the GNU C Library.
1527
1528 * The feature test macro __STDC_WANT_IEC_60559_FUNCS_EXT__, from ISO/IEC TS
1529 18661-4:2015, is supported to enable declarations of functions and macros
1530 from that TS. Note that most features from that TS are not supported by
1531 the GNU C Library.
1532
1533 * The nonstandard feature selection macros _REENTRANT and _THREAD_SAFE are
1534 now treated as compatibility synonyms for _POSIX_C_SOURCE=199506L.
1535 Since the GNU C Library defaults to a much newer revision of POSIX, this
1536 will only affect programs that specifically request an old conformance
1537 mode. For instance, a program compiled with -std=c89 -D_REENTRANT will
1538 see a change in the visible declarations, but a program compiled with
1539 just -D_REENTRANT, or -std=c99 -D_POSIX_C_SOURCE=200809L -D_REENTRANT,
1540 will not.
1541
1542 Some C libraries once required _REENTRANT and/or _THREAD_SAFE to be
1543 defined by all multithreaded code, but glibc has not required this for
1544 many years.
1545
1546 * The inclusion of <sys/sysmacros.h> by <sys/types.h> is deprecated. This
1547 means that in a future release, the macros “major”, “minor”, and “makedev”
1548 will only be available from <sys/sysmacros.h>.
1549
1550 These macros are not part of POSIX nor XSI, and their names frequently
1551 collide with user code; see for instance glibc bug 19239 and Red Hat bug
1552 130601. <stdlib.h> includes <sys/types.h> under _GNU_SOURCE, and C++ code
1553 presently cannot avoid being compiled under _GNU_SOURCE, exacerbating the
1554 problem.
1555
1556 * New <fenv.h> features from TS 18661-1:2014 are added to libm: the
1557 fesetexcept, fetestexceptflag, fegetmode and fesetmode functions, the
1558 femode_t type and the FE_DFL_MODE and FE_SNANS_ALWAYS_SIGNAL macros.
1559
1560 * Integer width macros from TS 18661-1:2014 are added to <limits.h>:
1561 CHAR_WIDTH, SCHAR_WIDTH, UCHAR_WIDTH, SHRT_WIDTH, USHRT_WIDTH, INT_WIDTH,
1562 UINT_WIDTH, LONG_WIDTH, ULONG_WIDTH, LLONG_WIDTH, ULLONG_WIDTH; and to
1563 <stdint.h>: INT8_WIDTH, UINT8_WIDTH, INT16_WIDTH, UINT16_WIDTH,
1564 INT32_WIDTH, UINT32_WIDTH, INT64_WIDTH, UINT64_WIDTH, INT_LEAST8_WIDTH,
1565 UINT_LEAST8_WIDTH, INT_LEAST16_WIDTH, UINT_LEAST16_WIDTH,
1566 INT_LEAST32_WIDTH, UINT_LEAST32_WIDTH, INT_LEAST64_WIDTH,
1567 UINT_LEAST64_WIDTH, INT_FAST8_WIDTH, UINT_FAST8_WIDTH, INT_FAST16_WIDTH,
1568 UINT_FAST16_WIDTH, INT_FAST32_WIDTH, UINT_FAST32_WIDTH, INT_FAST64_WIDTH,
1569 UINT_FAST64_WIDTH, INTPTR_WIDTH, UINTPTR_WIDTH, INTMAX_WIDTH,
1570 UINTMAX_WIDTH, PTRDIFF_WIDTH, SIG_ATOMIC_WIDTH, SIZE_WIDTH, WCHAR_WIDTH,
1571 WINT_WIDTH.
1572
1573 * New <math.h> features are added from TS 18661-1:2014:
1574
1575 - Signaling NaN macros: SNANF, SNAN, SNANL.
1576
1577 - Nearest integer functions: roundeven, roundevenf, roundevenl, fromfp,
1578 fromfpf, fromfpl, ufromfp, ufromfpf, ufromfpl, fromfpx, fromfpxf,
1579 fromfpxl, ufromfpx, ufromfpxf, ufromfpxl.
1580
1581 - llogb functions: the llogb, llogbf and llogbl functions, and the
1582 FP_LLOGB0 and FP_LLOGBNAN macros.
1583
1584 - Max-min magnitude functions: fmaxmag, fmaxmagf, fmaxmagl, fminmag,
1585 fminmagf, fminmagl.
1586
1587 - Comparison macros: iseqsig.
1588
1589 - Classification macros: iscanonical, issubnormal, iszero.
1590
1591 - Total order functions: totalorder, totalorderf, totalorderl,
1592 totalordermag, totalordermagf, totalordermagl.
1593
1594 - Canonicalize functions: canonicalize, canonicalizef, canonicalizel.
1595
1596 - NaN functions: getpayload, getpayloadf, getpayloadl, setpayload,
1597 setpayloadf, setpayloadl, setpayloadsig, setpayloadsigf, setpayloadsigl.
1598
1599 * The functions strfromd, strfromf, and strfroml, from ISO/IEC TS 18661-1:2014,
1600 are added to libc. They convert a floating-point number into string.
1601
1602 * Most of glibc can now be built with the stack smashing protector enabled.
1603 It is recommended to build glibc with --enable-stack-protector=strong.
1604 Implemented by Nick Alcock (Oracle).
1605
1606 * The function explicit_bzero, from OpenBSD, has been added to libc. It is
1607 intended to be used instead of memset() to erase sensitive data after use;
1608 the compiler will not optimize out calls to explicit_bzero even if they
1609 are "unnecessary" (in the sense that no _correct_ program can observe the
1610 effects of the memory clear).
1611
1612 * On ColdFire, MicroBlaze, Nios II and SH3, the float_t type is now defined
1613 to float instead of double. This does not affect the ABI of any libraries
1614 that are part of the GNU C Library, but may affect the ABI of other
1615 libraries that use this type in their interfaces.
1616
1617 * On x86_64, when compiling with -mfpmath=387 or -mfpmath=sse+387, the
1618 float_t and double_t types are now defined to long double instead of float
1619 and double. These options are not the default, and this does not affect
1620 the ABI of any libraries that are part of the GNU C Library, but it may
1621 affect the ABI of other libraries that use this type in their interfaces,
1622 if they are compiled or used with those options.
1623
1624 * The getentropy and getrandom functions, and the <sys/random.h> header file
1625 have been added.
1626
1627 * The buffer size for byte-oriented stdio streams is now limited to 8192
1628 bytes by default. Previously, on Linux, the default buffer size on most
1629 file systems was 4096 bytes (and thus remains unchanged), except on
1630 network file systems, where the buffer size was unpredictable and could be
1631 as large as several megabytes.
1632
1633 * The <sys/quota.h> header now includes the <linux/quota.h> header. Support
1634 for the Linux quota interface which predates kernel version 2.4.22 has
1635 been removed.
1636
1637 * The malloc_get_state and malloc_set_state functions have been removed.
1638 Already-existing binaries that dynamically link to these functions will
1639 get a hidden implementation in which malloc_get_state is a stub. As far
1640 as we know, these functions are used only by GNU Emacs and this change
1641 will not adversely affect already-built Emacs executables. Any undumped
1642 Emacs executables, which normally exist only during an Emacs build, should
1643 be rebuilt by re-running “./configure; make” in the Emacs build tree.
1644
1645 * The “ip6-dotint” and “no-ip6-dotint” resolver options, and the
1646 corresponding RES_NOIP6DOTINT flag from <resolv.h> have been removed.
1647 “no-ip6-dotint” had already been the default, and support for the
1648 “ip6-dotint” option was removed from the Internet in 2006.
1649
1650 * The "ip6-bytestring" resolver option and the corresponding RES_USEBSTRING
1651 flag from <resolv.h> have been removed. The option relied on a
1652 backwards-incompatible DNS extension which was never deployed on the
1653 Internet.
1654
1655 * The flags RES_AAONLY, RES_PRIMARY, RES_NOCHECKNAME, RES_KEEPTSIG,
1656 RES_BLAST defined in the <resolv.h> header file have been deprecated.
1657 They were already unimplemented.
1658
1659 * The "inet6" option in /etc/resolv.conf and the RES_USE_INET6 flag for
1660 _res.flags are deprecated. The flag was standardized in RFC 2133, but
1661 removed again from the IETF name lookup interface specification in RFC
1662 2553. Applications should use getaddrinfo instead.
1663
1664 * DNSSEC-related declarations and definitions have been removed from the
1665 <arpa/nameser.h> header file, and libresolv will no longer attempt to
1666 decode the data part of DNSSEC record types. Previous versions of glibc
1667 only implemented minimal support for the previous version of DNSSEC, which
1668 is incompatible with the currently deployed version.
1669
1670 * The resource record type classification macros ns_t_qt_p, ns_t_mrr_p,
1671 ns_t_rr_p, ns_t_udp_p, ns_t_xfr_p have been removed from the
1672 <arpa/nameser.h> header file because the distinction between RR types and
1673 meta-RR types is not officially standardized, subject to revision, and
1674 thus not suitable for encoding in a macro.
1675
1676 * The types res_sendhookact, res_send_qhook, re_send_rhook, and the qhook
1677 and rhook members of the res_state type in <resolv.h> have been removed.
1678 The glibc stub resolver did not support these hooks, but the header file
1679 did not reflect that.
1680
1681 * For multi-arch support it is recommended to use a GCC which has
1682 been built with support for GNU indirect functions. This ensures
1683 that correct debugging information is generated for functions
1684 selected by IFUNC resolvers. This support can either be enabled by
1685 configuring GCC with '--enable-gnu-indirect-function', or by
1686 enabling it by default by setting 'default_gnu_indirect_function'
1687 variable for a particular architecture in the GCC source file
1688 'gcc/config.gcc'.
1689
1690 * GDB pretty printers have been added for mutex and condition variable
1691 structures in POSIX Threads. When installed and loaded in gdb these pretty
1692 printers show various pthread variables in human-readable form when read
1693 using the 'print' or 'display' commands in gdb.
1694
1695 * Tunables feature added to allow tweaking of the runtime for an application
1696 program. This feature can be enabled with the '--enable-tunables' configure
1697 flag. The GNU C Library manual has details on usage and README.tunables has
1698 instructions on adding new tunables to the library.
1699
1700 * A new version of condition variables functions have been implemented in
1701 the NPTL implementation of POSIX Threads to provide stronger ordering
1702 guarantees.
1703
1704 * A new version of pthread_rwlock functions have been implemented to use a more
1705 scalable algorithm primarily through not using a critical section anymore to
1706 make state changes.
1707
1708 Security related changes:
1709
1710 * On ARM EABI (32-bit), generating a backtrace for execution contexts which
1711 have been created with makecontext could fail to terminate due to a
1712 missing .cantunwind annotation. This has been observed to lead to a hang
1713 (denial of service) in some Go applications compiled with gccgo. Reported
1714 by Andreas Schwab. (CVE-2016-6323)
1715
1716 * The DNS stub resolver functions would crash due to a NULL pointer
1717 dereference when processing a query with a valid DNS question type which
1718 was used internally in the implementation. The stub resolver now uses a
1719 question type which is outside the range of valid question type values.
1720 (CVE-2015-5180)
1721
1722 The following bugs are resolved with this release:
1723
1724 [4099] stdio: Overly agressive caching by stream i/o functions.
1725 [7065] build: Support building glibc with -fstack-protector or -fstack-
1726 protector-all
1727 [9842] localedata: en_CA: incorrect date format
1728 [13165] nptl: pthread_cond_wait() can consume a signal that was sent
1729 before it started waiting
1730 [14139] manual: Do not hardcode platform names in manual/libm-err-tab.pl
1731 [15765] nptl: sem_open is wrongly a cancellation point
1732 [16421] network: IN6_IS_ADDR_UNSPECIFIED can use undefined s6_addr32
1733 [16458] libc: endian.h and netinet/in.h byte order macros return values of
1734 wrong type
1735 [16628] dynamic-link: Segfault after a binary without pthread dlopen()s a
1736 library linked with pthread
1737 [16630] nptl: Use SYSENTER for pthread_cond_broadcast/signal() (i.e. fix
1738 "FIXME: Ingo" issue)
1739 [16907] libc: <argp.h> compiled with --std=cXX disables __attribute__
1740 [17252] libc: getrandom and getentropy syscall
1741 [17730] malloc: thread-local storage is sometimes improperly free()'d
1742 after being __libc_memalign()'d
1743 [18241] stdio: failed fseek on memstream does not set errno and can fail
1744 when it shouldnt
1745 [18243] nptl: sem_wait, sem_timedwait are cancellation points shm_open is
1746 not
1747 [18463] nptl: pthread_cond_broadcast issue when surrounded by
1748 PTHREAD_PRIO_INHERIT mutex on ARM
1749 [18784] network: res_query and related function crash for special record
1750 type queries (CVE-2015-5180)
1751 [19380] math: strtod does not raise "inexact"
1752 [19387] string: Integer overflow in memchr
1753 [19390] string: Integer overflow in strncat
1754 [19398] build: linknamespace tests fail with massively parallel build
1755 [19402] nptl: Deadlock with robust shared mutex and asynchronous
1756 termination
1757 [19469] malloc: M_PERTURB in test-skeleton.c invalidates malloc tests
1758 [19473] malloc: Turn malloc_get_state etc. in compatibility symbols
1759 [19514] libc: [PATCH] Fix spelling errors in spelling
1760 "implement"/"implementation" in several places
1761 [19582] network: Deprecate RES_USE_INET6
1762 [19673] manual: clog10 docs appear to be erroneous
1763 [19810] dynamic-link: dlopen with both RTLD_NOLOAD and RTLD_NODELETE
1764 causes a segmentation fault
1765 [19826] libc: invalid pointer returned from __tls_get_addr with static
1766 linking
1767 [20016] network: resolv: Remove hooks support from the API
1768 [20019] dynamic-link: NULL pointer dereference in libc.so.6 IFUNC due to
1769 uninitialized GOT
1770 [20033] math: [x86_64] vectorized math function don't call the __finite
1771 versions
1772 [20116] nptl: use after free in pthread_create
1773 [20181] stdio: open_memstream(): writes not at end of stream corrupt data
1774 [20292] dynamic-link: Comparison in elf/dl-open.c _dl_addr_inside_object
1775 is always true.
1776 [20311] nptl: please install proc_service.h
1777 [20366] libc: Compilation errors in installed headers in strict-compliance
1778 modes
1779 [20370] malloc: malloc: Arena free list management is still racy
1780 (incorrect fix in bug 19243)
1781 [20386] libc: assert (X = 0) does not result in GCC warning
1782 [20432] malloc: malloc: Minimize interface required for interposition
1783 [20435] libc: Missing unwind info in __startcontext causes infinite loop
1784 in _Unwind_Backtrace (CVE-2016-6323)
1785 [20444] hurd: recvmsg: PF_LOCAL sockets and msg_name lead to SIGLOST
1786 [20452] nptl: Addition of sysdep.o to libpthread.a breaks relinking
1787 libpthread.a
1788 [20455] math: [powerpc] fesetexceptflag fails to clear FE_INVALID
1789 [20459] localedata: et_EE: locale has wrong {p,n}_cs_precedes value
1790 [20477] network: resolv: incorrect double-checked locking related to
1791 _res_hconf
1792 [20478] libc: libc_ifunc macro and similar usages leads to false debug-
1793 information.
1794 [20495] math: x86_64 performance degradation due to AVX/SSE transition
1795 penalty
1796 [20497] localedata: lt_LT: LC_TIME d_fmt used is obsolete
1797 [20508] dynamic-link: _dl_runtime_resolve_avx/_dl_runtime_profile_avx512
1798 cause transition penalty
1799 [20517] math: sparcv9 missing fdiml compat symbol
1800 [20524] manual: strverscmp is inconsistent
1801 [20525] libc: <sys/quota.h> should be based on kernel headers
1802 [20539] math: GCC 7 -static -lm fails to link at -O3
1803 [20554] libc: ld: bss-plt forced due to /usr/lib/libc_nonshared.a(ppc-
1804 mcount.oS)
1805 [20558] string: POSIX bcopy/bzero decls do not implement Object Size
1806 Checking
1807 [20591] network: Remove obsolete DNSSEC support
1808 [20592] network: DNS resource record type classification macros in
1809 <arpa/nameser.h> are incorrect
1810 [20593] network: Update DNS RR type definitions
1811 [20611] network: getaddrinfo accepts invalid numeric scope IDs
1812 [20615] build: glibc build fails when using --with-cpu=power9 --enable-
1813 multi-arch
1814 [20629] network: libresolv: Remove support for bitstring labels
1815 (RES_USEBSTRING)
1816 [20647] libc: GLIBC quitting every program - glibc on Pentium-S leads to
1817 assertion: "maxidx >= 2"
1818 [20660] math: [arm] Use VSQRT
1819 [20662] libc: checking whether x86_64-pc-linux-gnu-gcc implicitly enables
1820 -fstack-protector no (32bit gcc 6.2.0 pie and ssp enable)
1821 [20689] libc: Test for FMA should also check for AVX.
1822 [20707] glob: gl_pathv entries not set to NULL with GLOB_DOOFFS
1823 [20715] math: iszero macro breaks existing code
1824 [20718] math: [powerpc] copysignl raises "invalid" for sNaN
1825 [20728] libc: powerpc: Missing TOC stub in clone
1826 [20729] build: glibc-2.24 fails to build for i486 with -Os
1827 [20750] build: Build fails with default PIE enabled gcc-6.x
1828 [20768] math: [alpha] sqrt fegetenv namespace
1829 [20785] libc: binutils 2.28 fails to assemble power6/memset.S file in
1830 glibc
1831 [20787] math: float_t is defined as float incorrectly on x86_64 even if
1832 __FLT_EVAL_METHOD__ is 2
1833 [20790] network: rpcgen buffer overrun in get_prog_declaration
1834 [20822] nptl: race condition in __lll_unlock_elision on powerpc
1835 [20829] libc: crypt snprintf namespace
1836 [20847] libc: tst-vfork3 failure
1837 [20855] math: Default bits/mathdef.h has inappropriate float_t
1838 [20859] math: [sh4] FP_ILOGB0 invalid
1839 [20864] localedata: iconv: cp936 missing single-byte euro sign (0x80,
1840 U+20AC), not same as GBK
1841 [20915] dynamic-link: global-dynamic TLS broken on aarch64
1842 [20916] math: pow handling of sNaN arguments
1843 [20918] build: Building with --enable-nss-crypt fails tst-linkall-static.
1844 [20919] math: Bad pow (qNaN, 0) result with -lieee
1845 [20940] math: hypot sNaN handling
1846 [20947] math: fmax, fmin sNaN handling
1847 [20956] libc: debug/tst-backtrace3-6 don't work with -O3 anymore
1848 [20964] network: sunrpc: Stack-based buffer overflow in getrpcport with
1849 RES_USE_INET6
1850 [20971] string: powerpc64/power7 memchr overflows internal pointer check
1851 [20973] nptl: robust mutexes: Lost wake-ups
1852 [20974] locale: bs_BA: yesexpr/noexpr regexes accept any character
1853 [20978] nis: strlen on null pointer in nss_nisplus
1854 [20985] nptl: robust mutexes: lowlevelrobustlock assembly on x86 blocks on
1855 wrong condition
1856 [21014] string: i686 memchr overflows internal pointer check
1857 [21019] libc: [mips] n32 lseek incorrect on overflow
1858 [21022] libc: [microblaze] __backtrace get_frame_size namespace
1859 [21026] libc: [MIPS] readahead syscall is broken on n64
1860 [21028] math: Fallback fesetexceptflag should always succeed
1861 [21045] libc: [powerpc-nofpu] swapcontext does not restore signal mask
1862 [21047] math: arm: fpu_control.h: _FPU_GETCW/_FPU_SETCW is rejected by
1863 clang
1864 [21053] libc: [SH] Namespace pollution from sys/ucontext.h
1865 [21061] librt: [microblaze] librt lost clock_* exports
1866 [21073] libc: tunables: insecure environment variables passed to
1867 subprocesses with AT_SECURE
1868 [21081] string: Missing vzeroupper in memset-vec-unaligned-erms.S
1869 \f
1870 Version 2.24
1871
1872 * The minimum Linux kernel version that this version of the GNU C Library
1873 can be used with is 3.2, except on i[4567]86 and x86_64, where Linux
1874 kernel version 2.6.32 or later suffices (on architectures that already
1875 required kernel versions more recent than 3.2, those requirements remain
1876 unchanged). Linux 3.2 or later kernel headers are required on all
1877 architectures.
1878
1879 * The pap_AN locale has been deleted. This has been deprecated for a long
1880 time. It has been replaced by pap_AW & pap_CW, both of which have long
1881 been included in previous releases.
1882
1883 * The readdir_r and readdir64_r functions have been deprecated. It is
1884 recommended to use readdir and readdir64 instead.
1885
1886 * The type “union wait” has been removed. It was deprecated in the early
1887 1990s and never part of POSIX. Application code should use the int type
1888 instead of “union wait”.
1889
1890 * A new NSS action is added to facilitate large distributed system
1891 administration. The action, MERGE, allows remote user stores like LDAP
1892 to be merged into local user stores like /etc/groups in order to provide
1893 easy to use, updated, and managed sets of merged credentials. The new
1894 action can be used by configuring it in /etc/nsswitch.conf:
1895 group: files [SUCCESS=merge] nis
1896 Implemented by Stephen Gallagher (Red Hat).
1897
1898 * The deprecated __malloc_initialize_hook variable has been removed from the
1899 API.
1900
1901 * The long unused localedef --old-style option has been removed. It hasn't
1902 done anything in over 16 years. Scripts using this option can safely
1903 drop it.
1904
1905 * nextupl, nextup, nextupf, nextdownl, nextdown and nextdownf are added to
1906 libm. They are defined by TS 18661 and IEEE754-2008. The nextup functions
1907 return the next representable value in the direction of positive infinity
1908 and the nextdown functions return the next representable value in the
1909 direction of negative infinity. These are currently enabled as GNU
1910 extensions.
1911
1912 Security related changes:
1913
1914 * An unnecessary stack copy in _nss_dns_getnetbyname_r was removed. It
1915 could result in a stack overflow when getnetbyname was called with an
1916 overly long name. (CVE-2016-3075)
1917
1918 * Previously, getaddrinfo copied large amounts of address data to the stack,
1919 even after the fix for CVE-2013-4458 has been applied, potentially
1920 resulting in a stack overflow. getaddrinfo now uses a heap allocation
1921 instead. Reported by Michael Petlan. (CVE-2016-3706)
1922
1923 * The glob function suffered from a stack-based buffer overflow when it was
1924 called with the GLOB_ALTDIRFUNC flag and encountered a long file name.
1925 Reported by Alexander Cherepanov. (CVE-2016-1234)
1926
1927 * The Sun RPC UDP client could exhaust all available stack space when
1928 flooded with crafted ICMP and UDP messages. Reported by Aldy Hernandez'
1929 alloca plugin for GCC. (CVE-2016-4429)
1930
1931 * The IPv6 name server management code in libresolv could result in a memory
1932 leak for each thread which is created, performs a failing naming lookup,
1933 and exits. Over time, this could result in a denial of service due to
1934 memory exhaustion. Reported by Matthias Schiffer. (CVE-2016-5417)
1935
1936 The following bugs are resolved with this release:
1937
1938 [1170] localedata: ne_NP: update Nepali locale definition file
1939 [3629] manual: stpcpy description in string.texi refers to MS-DOG instead
1940 of MS-DOS.
1941 [6527] malloc: [powerpc] Malloc alignment insufficient for PowerPC
1942 [6796] math: fdim() does not set errno on overflow
1943 [10354] libc: posix_spawn should use vfork() in more cases than presently
1944 [11213] localedata: localedata: add copyright disclaimer to locale files
1945 [12143] localedata: chr_US: new Cherokee locale
1946 [12450] localedata: sgs_LT: new locale
1947 [12676] localedata: ln_CD: new locale
1948 [13237] localedata: LC_ADDRESS.country_name: update all locales w/latest
1949 CLDR data
1950 [13304] math: fma, fmaf, fmal produce wrong results
1951 [14259] build: --localedir arg to configure is ignored
1952 [14499] nptl: Does posix_spawn invoke atfork handlers / use vfork?
1953 [14750] libc: Race condition in posix_spawn vfork usage vs signal handlers
1954 [14934] localedata: es_CL: wrong first weekday chilean locale
1955 [15262] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of
1956 romanisation
1957 [15263] localedata: LC_MESSAGES.yesexpr/noexpr: inconsistent use of 1/0
1958 and +/-
1959 [15264] localedata: LC_MESSAGES.yesstr/nostr: lacking in many locales
1960 [15368] nptl: raise() is not async-signal-safe
1961 [15479] math: ceil, floor, round and trunc raise inexact exception
1962 [15578] localedata: kk_KZ: various updates
1963 [16003] localedata: pap_AN: punt old locale
1964 [16137] localedata: iw_IL: punt old locale
1965 [16190] localedata: eo: new esperanto locale
1966 [16374] localedata: lv_LV: change currency symbol in LC_MONETARY to euro
1967 [16742] malloc: race condition: pthread_atfork() called before first
1968 malloc() results in unexpected locking behaviour/deadlocks
1969 [16975] localedata: LC_MESSAGES.yesexpr/noexpr: revisit capitalization in
1970 all locales
1971 [16983] localedata: postal_fmt does not allow %l and %n modifiers
1972 [17565] localedata: pt_PT: wrong (work-)week start
1973 [17899] math: [powerpc] floorl returns negative zero with FE_DOWNWARD
1974 [17950] build: Build fails with -msse
1975 [18205] localedata: be_BY*: wrong first_weekday and first_workday
1976 [18433] libc: posix_spawn does not return correctly upon failure to
1977 execute
1978 [18453] localedata: charmaps/IBM875: incorrect codes
1979 [18712] string: bits/string2.h incompatible with -O2 -Werror=packed
1980 -Wsystem-headers
1981 [18896] localedata: he_IL: improvements for currency
1982 [18911] localedata: ro_RO: Correcting week day name for "Tuesday" in
1983 Romanian locale data
1984 [18960] locale: s390: _nl_locale_subfreeres uses larl opcode on misaligned
1985 symbol
1986 [19056] libc: Deprecate readdir_r
1987 [19133] localedata: pt_*: days & months should be lowercase in Portuguese
1988 language
1989 [19198] localedata: nl_NL: small improvements for Dutch locales
1990 [19257] network: Per-thread memory leak in __res_vinit with IPv6
1991 nameservers (CVE-2016-5417)
1992 [19269] build: tst-audit4 and tst-audit10 failures with gcc-6 on non avx
1993 machine
1994 [19400] locale: Language missing in "iso-639.def", trivial fix in
1995 description
1996 [19431] malloc: Deadlock between fflush, getdelim, and fork
1997 [19505] libc: Incorrect file descriptor validity checks in
1998 posix_spawn_file_actions_add{open,close,dup2}
1999 [19509] dynamic-link: dlsym, dlvsym do not report errors through dlerror
2000 when using RTLD_NEXT
2001 [19512] locale: Stale `#ifndef HAVE_BUILTIN_EXPECT' in
2002 `intl/{gettextP,loadinfo}.h'
2003 [19534] libc: execle, execlp may use malloc
2004 [19568] localedata: *_CH: Swiss locales have inconsistent start of week
2005 [19573] network: res_nclose and __res_maybe_init disagree about name
2006 server initialization, breaking Hesiod
2007 [19575] localedata: Status of GB18030 tables
2008 [19581] localedata: sr_* date_fmt string contains additional newline
2009 [19583] string: SSSE3_Fast_Copy_Backward flag needs to be enabled for AMD
2010 Excavator core
2011 [19592] math: [ldbl-128ibm] ceill incorrect in non-default rounding modes
2012 [19593] math: [ldbl-128ibm] truncl incorrect in non-default rounding modes
2013 [19594] math: [ldbl-128ibm] roundl incorrect in non-default rounding modes
2014 [19595] math: [ldbl-128ibm] fmodl incorrect for results in subnormal
2015 double range
2016 [19602] math: [ldbl-128ibm] fmodl handling of equal arguments with low
2017 part zero incorrect
2018 [19603] math: [ldbl-128ibm] remainderl, remquol incorrect sign handling in
2019 equality tests
2020 [19610] dynamic-link: ldconfig -X removes stale symbolic links
2021 [19613] libc: s390x (64 bit) macro expansion WCOREDUMP and others
2022 [19633] locale: strfmon_l applies global locale to number formatting
2023 [19642] network: Memory leak in getnameinfo
2024 [19648] libc: test-skeleton.c: Do not set RLIMIT_DATA
2025 [19653] libc: Potential for NULL pointer dereference (CWE-476) in
2026 glibc-2.22
2027 [19654] math: [x86_64] Need testcase for BZ #19590 fix
2028 [19671] localedata: Missing Sanity Check for malloc() in 'tst-fmon.c' &
2029 'tst-numeric.c'
2030 [19674] math: [ldbl-128ibm] powl incorrect overflow handling
2031 [19677] math: [ldbl-128ibm] remainderl equality test incorrect for zero
2032 low part
2033 [19678] math: [ldbl-128ibm] nextafterl, nexttowardl incorrect sign of zero
2034 result
2035 [19679] dynamic-link: gcc-4.9.3 C++ exception handling broken due to
2036 unaligned stack
2037 [19726] locale: Converting UCS4LE to INTERNAL with iconv() does not update
2038 pointers and lengths in error-case.
2039 [19727] locale: Converting from/to UTF-xx with iconv() does not always
2040 report errors on UTF-16 surrogates values.
2041 [19755] nscd: nscd assertion failure in gc
2042 [19758] dynamic-link: Typo in EXTRA_LD_ENVVARS for x86-64
2043 [19759] libc: mempcpy shouldn't be inlined
2044 [19762] dynamic-link: HAS_CPU_FEATURE/HAS_ARCH_FEATURE are easy to misuse
2045 [19765] libc: s390 needs an optimized mempcpy
2046 [19779] glob: glob: buffer overflow with GLOB_ALTDIRFUNC due to incorrect
2047 NAME_MAX limit assumption (CVE-2016-1234)
2048 [19783] build: benchtests don't support --enable-hardcoded-path-in-tests
2049 [19787] network: Missing and incorrect truncation checks in getnameinfo
2050 [19790] math: [ldbl-128ibm] nearbyintl incorrect in non-default rounding
2051 modes
2052 [19791] network: Assertion failure in res_query.c with un-connectable name
2053 server addresses
2054 [19792] libc: MIPS: backtrace yields infinite backtrace with makecontext
2055 [19822] math: libm.so install clobbers old version
2056 [19825] network: resolv: send_vc can return uninitialized data in second
2057 response to getaddrinfo
2058 [19830] network: nss_dns: should check RDATA length against buffer length
2059 [19831] network: nss_dns: getaddrinfo returns uninitialized data when
2060 confronted with A/AAAA records of invalid size
2061 [19837] nss: nss_db: No retries for some long lines with a larger buffer
2062 [19848] math: powl(10,n) for n=-4,-5,-6,-7 is off by more than 1 ULP
2063 [19853] stdio: Printing IBM long double in decimal with high precision is
2064 sometimes incorrect
2065 [19860] build: x86_64: compile errors for tst-audit10 and tst-auditmod10b
2066 [19861] nptl: libpthread IFUNC resolver for fork can lead to crash
2067 [19862] network: resolv, nss_dns: Remove remaining logging of unexpected
2068 record types
2069 [19865] network: Assertion failure or memory leak in
2070 _nss_dns_getcanonname_r
2071 [19868] network: nss_dns: netent code does not skip over non-PTR records
2072 [19879] network: nss_dns: Stack overflow in getnetbyname implementation
2073 (CVE-2016-3075)
2074 [19881] string: Improve x86-64 memset
2075 [19907] string: Incorrect memcpy tests
2076 [19916] dynamic-link: S390: fprs/vrs are not saved/restored while
2077 resolving symbols
2078 [19925] libc: termios.h XCASE namespace
2079 [19928] string: memmove-vec-unaligned-erms.S is slow with large data size
2080 [19929] libc: limits.h NL_NMAX namespace
2081 [19931] stdio: Memory leak in vfprintf
2082 [19957] libc: clone(CLONE_VM) access invalid parent memory
2083 [19963] localedata: en_IL: New locale
2084 [19989] stdio: stdio.h cuserid namespace
2085 [19994] network: getaddrinfo does not restore RES_USE_INET6 flag in
2086 gethosts
2087 [19996] locale: langinfo.h nl_langinfo_l namespace
2088 [20005] stdio: fflush on a file opened with fmemopen resets position to 0
2089 [20010] network: getaddrinfo: Stack overflow in hostent translation
2090 (CVE-2016-3706)
2091 [20012] stdio: libio: fmemopen append mode failure
2092 [20014] stdio: stdio.h namespace for pre-threads POSIX
2093 [20017] network: resolv: Use gmtime_r instead of gmtime in p_secstodate
2094 [20023] libc: fcntl.h timespec namespace
2095 [20024] math: [x86_64] vectorized sincos trashes the stack
2096 [20031] network: nss_hesiod: Heap overflow in get_txt_records
2097 [20041] time: sys/time.h timespec namespace
2098 [20043] libc: unistd.h missing cuserid for UNIX98 and before
2099 [20044] libc: unistd.h missing pthread_atfork for UNIX98
2100 [20051] libc: ttyslot in wrong header under wrong conditions
2101 [20054] libc: gethostname not declared for XPG4
2102 [20055] libc: termios.h missing tcgetsid for XPG4
2103 [20072] dynamic-link: x86 init_cpu_features is called twice in static
2104 executable
2105 [20073] libc: sys/stat.h fchmod namespace
2106 [20074] libc: stdlib.h rand_r namespace
2107 [20076] libc: sys/stat.h missing S_IFSOCK, S_ISSOCK for XPG4
2108 [20094] libc: stdlib.h should not declare grantpt, ptsname, unlockpt for
2109 XPG3
2110 [20111] libc: struct sockaddr_storage cannot be aggregate-copied
2111 [20112] network: sunrpc: stack (frame) overflow in Sun RPC clntudp_call
2112 (CVE-2016-4429)
2113 [20115] string: Extra alignment in memset-vec-unaligned-erms.S
2114 [20119] libc: Wrong mask for processors level type from CPUID
2115 [20139] dynamic-link: Upper part of zmm is zeroed if Glibc is built with
2116 AS not supporting AVX512
2117 [20151] math: [ldbl-128/ldbl-128ibm] j0l, j1l, y0l, y1l return sNaN for
2118 sNaN argument
2119 [20153] math: [ldbl-128ibm] sqrtl (sNaN) returns sNaN
2120 [20156] math: [ldbl-128ibm] ceill, rintl etc. return sNaN for sNaN
2121 argument
2122 [20157] math: [powerpc] fabsl (sNaN) wrongly raises "invalid"
2123 [20160] math: [powerpc] ceil, rint etc. return sNaN for sNaN input
2124 [20178] libc: posix_spawn{p} should not call exit
2125 [20191] stdio: libio: vtables hardening
2126 [20195] string: FMA4 detection requires CPUID execution with register
2127 eax=0x80000001
2128 [20198] libc: quick_exit incorrectly destroys C++11 thread objects.
2129 [20205] math: [i386/x86_64] nextafterl incorrect incrementing negative
2130 subnormals
2131 [20212] math: acos (sNaN) returns sNaN
2132 [20213] math: asin (sNaN) returns sNaN
2133 [20214] network: Linux header sync with linux/in6.h and ipv6.h again.
2134 [20218] math: [i386] asinhl (sNaN) returns sNaN
2135 [20219] math: [i386] atanhl (sNaN) returns sNaN
2136 [20222] stdio: fopencookie: Mangle function pointers
2137 [20224] math: [i386] cbrtl (sNaN) returns sNaN
2138 [20225] math: ldexp, scalbn, scalbln return sNaN for sNaN input
2139 [20226] math: [i386/x86_64] expl, exp10l, expm1l return sNaN for sNaN
2140 input
2141 [20227] math: [i386/x86_64] logl (sNaN) returns sNaN
2142 [20228] math: [i386/x86_64] log10l (sNaN) returns sNaN
2143 [20229] math: [i386/x86_64] log1pl (sNaN) returns sNaN
2144 [20232] math: [ldbl-128] expm1l (sNaN) returns sNaN
2145 [20233] math: [ldbl-128ibm] expm1l (sNaN) returns sNaN
2146 [20234] math: [ldbl-128ibm] log1pl (sNaN) returns sNaN
2147 [20235] math: [i386/x86_64] log2l (sNaN) returns sNaN
2148 [20237] nss: nss_db: get*ent segfaults without preceding set*ent
2149 [20240] math: modf (sNaN) returns sNaN
2150 [20248] libc: debug/tst-longjump_chk2 calls printf from a signal handler
2151 [20250] math: frexp (sNaN) returns sNaN
2152 [20252] math: atan2 (sNaN, qNaN) fails to raise "invalid"
2153 [20255] math: [i386] fdim, fdimf return with excess range and precision /
2154 double rounding
2155 [20256] math: [i386/x86_64] fdiml returns sNaN for sNaN input
2156 [20260] string: ../sysdeps/x86/bits/string.h:1092:3: error: array
2157 subscript is below array bounds [-Werror=array-bounds]
2158 [20262] nis: _nss_nis_initgroups_dyn always returns NSS_STATUS_NOTFOUND
2159 [20263] nptl: robust mutex deadlocks if other thread requests timedlock
2160 (Only arm/linux)
2161 [20277] libc: $dp is not initialized correctly in sysdeps/hppa/start.S
2162 [20284] malloc: malloc: Corrupt arena avoidance causes unnecessary mmap
2163 fallbacks
2164 [20296] math: [i386/x86_64] scalbl returns sNaN for sNaN input, missing
2165 "invalid" exceptions
2166 [20314] nptl: make[4]: *** [/usr/include/stdlib.h] Error 1
2167 [20316] localedata: id_ID: Februari instead of Pebruari
2168 [20327] string: POWER8 strcasecmp returns incorrect result
2169 [20347] math: Failure: Test: j0_downward (0xap+0)
2170 [20348] libc: FAIL: misc/tst-preadvwritev64
2171 [20349] libc: 64-bit value is passed differently in p{readv,writev}{64}
2172 [20350] libc: There is no test for p{read,write}64
2173 [20357] math: Incorrect cos result for 1.5174239687223976
2174 [20384] build: Don't run libmvec-sincos-avx* tests on non avx machines
2175 \f
2176 Version 2.23
2177
2178 * Unicode 8.0.0 Support: Character encoding, character type info, and
2179 transliteration tables are all updated to Unicode 8.0.0, using new
2180 and/or improved generator scripts contributed by Mike FABIAN (Red Hat).
2181 These updates cause user visible changes, such as the fixes for bugs
2182 89, 16061, and 18568.
2183
2184 * sched_setaffinity, pthread_setaffinity_np no longer attempt to guess the
2185 kernel-internal CPU set size. This means that requests that change the
2186 CPU affinity which failed before (for example, an all-ones CPU mask) will
2187 now succeed. Applications that need to determine the effective CPU
2188 affinities need to call sched_getaffinity or pthread_getaffinity_np after
2189 setting it because the kernel can adjust it (and the previous size check
2190 would not detect this in the majority of cases).
2191
2192 * The fts.h header can now be used with -D_FILE_OFFSET_BITS=64. With LFS
2193 the following new symbols are used: fts64_children, fts64_close,
2194 fts64_open, fts64_read and fts64_set.
2195
2196 * getaddrinfo now detects certain invalid responses on an internal netlink
2197 socket. If such responses are received, an affected process will
2198 terminate with an error message of "Unexpected error <number> on netlink
2199 descriptor <number>" or "Unexpected netlink response of size <number> on
2200 descriptor <number>". The most likely cause for these errors is a
2201 multi-threaded application which erroneously closes and reuses the netlink
2202 file descriptor while it is used by getaddrinfo.
2203
2204 * A defect in the malloc implementation, present since glibc 2.15 (2012) or
2205 glibc 2.10 via --enable-experimental-malloc (2009), could result in the
2206 unnecessary serialization of memory allocation requests across threads.
2207 The defect is now corrected. Users should see a substantial increase in
2208 the concurent throughput of allocation requests for applications which
2209 trigger this bug. Affected applications typically create and
2210 destroy threads frequently. (Bug 19048 was reported and analyzed by
2211 Ericsson.)
2212
2213 * There is now a --disable-timezone-tools configure option for disabling the
2214 building and installing of the timezone related utilities (zic, zdump, and
2215 tzselect). This is useful for people who build the timezone data and code
2216 independent of the GNU C Library.
2217
2218 * The obsolete header <regexp.h> has been removed. Programs that require
2219 this header must be updated to use <regex.h> instead.
2220
2221 * The obsolete functions bdflush, create_module, get_kernel_syms,
2222 query_module and uselib are no longer available to newly linked binaries;
2223 the header <sys/kdaemon.h> has been removed. These functions and header
2224 were specific to systems using the Linux kernel and could not usefully be
2225 used with the GNU C Library on systems with version 2.6 or later of the
2226 Linux kernel.
2227
2228 * Optimized string, wcsmbs and memory functions for IBM z13.
2229 Implemented by Stefan Liebler.
2230
2231 * Newly linked programs that define a variable called signgam will no longer
2232 have it set by the lgamma, lgammaf and lgammal functions. Programs that
2233 require signgam to be set by those functions must ensure that they use the
2234 variable provided by the GNU C Library and declared in <math.h>, without
2235 defining their own copy.
2236
2237 * The minimum GCC version that can be used to build this version of the GNU
2238 C Library is GCC 4.7. Older GCC versions, and non-GNU compilers, can
2239 still be used to compile programs using the GNU C Library.
2240
2241 Security related changes:
2242
2243 * An out-of-bounds value in a broken-out struct tm argument to strftime no
2244 longer causes a crash. Reported by Adam Nielsen. (CVE-2015-8776)
2245
2246 * The LD_POINTER_GUARD environment variable can no longer be used to disable
2247 the pointer guard feature. It is always enabled. Previously,
2248 LD_POINTER_GUARD could be used to disable security hardening in binaries
2249 running in privileged AT_SECURE mode. Reported by Hector Marco-Gisbert.
2250 (CVE-2015-8777)
2251
2252 * An integer overflow in hcreate and hcreate_r could lead to an
2253 out-of-bounds memory access. Reported by Szabolcs Nagy. (CVE-2015-8778)
2254
2255 * The catopen function no longer has unbounded stack usage. Reported by
2256 Max. (CVE-2015-8779)
2257
2258 * The nan, nanf and nanl functions no longer have unbounded stack usage
2259 depending on the length of the string passed as an argument to the
2260 functions. Reported by Joseph Myers. (CVE-2014-9761)
2261
2262 * A stack-based buffer overflow was found in libresolv when invoked from
2263 libnss_dns, allowing specially crafted DNS responses to seize control
2264 of execution flow in the DNS client. The buffer overflow occurs in
2265 the functions send_dg (send datagram) and send_vc (send TCP) for the
2266 NSS module libnss_dns.so.2 when calling getaddrinfo with AF_UNSPEC
2267 family. The use of AF_UNSPEC triggers the low-level resolver code to
2268 send out two parallel queries for A and AAAA. A mismanagement of the
2269 buffers used for those queries could result in the response of a query
2270 writing beyond the alloca allocated buffer created by
2271 _nss_dns_gethostbyname4_r. Buffer management is simplified to remove
2272 the overflow. Thanks to the Google Security Team and Red Hat for
2273 reporting the security impact of this issue, and Robert Holiday of
2274 Ciena for reporting the related bug 18665. (CVE-2015-7547)
2275
2276 The following bugs are resolved with this release:
2277
2278 [89] localedata: Locales nb_NO and nn_NO should transliterate æøå
2279 [887] math: Math library function "logb" and "nextafter" inconsistent
2280 [2542] math: Incorrect return from float gamma (-0X1.FA471547C2FE5P+1)
2281 [2543] math: Incorrect return from float gamma (-0X1.9260DCP+1)
2282 [2558] math: Incorrect return from double gamma (-0X1.FA471547C2FE5P+1)
2283 [2898] libc: [improve] warning: the use of `mktemp' is dangerous, better
2284 use `mkstemp'
2285 [4404] localedata: German translation of "Alarm clock" is misleading
2286 [6799] math: nextafter() and nexttoward() doen't set errno on
2287 overflow/underflow errors
2288 [6803] math: scalb(), scalbln(), scalbn() do not set errno on
2289 overflow/underflow
2290 [10432] nis: _nss_nis_setnetgrent assertion failure
2291 [11460] libc: fts has no LFS support
2292 [12926] network: getaddrinfo()/make_request() may spin forever
2293 [13065] nptl: Race condition in pthread barriers
2294 [13690] nptl: pthread_mutex_unlock potentially cause invalid access
2295 [14341] dynamic-link: Dynamic linker crash when DT_JMPREL and DT_REL{,A}
2296 are not contiguous
2297 [14551] math: [ldbl-128ibm] strtold overflow handling for IBM long double
2298 [14912] libc: Rename non-installed bits/*.h headers
2299 [15002] libc: Avoid undefined behavior in posix_fallocate overflow check
2300 [15367] math: Let gcc use __builtin_isinf
2301 [15384] math: One constant fewer in ieee754/dbl-64/wordsize-64/s_finite.c
2302 [15421] math: lgamma wrongly sets signgam for ISO C
2303 [15470] math: [arm] On ARM llrintl() and llroundl() do not raise
2304 FE_INVALID with argument out of range
2305 [15491] math: [i386/x86_64] x86 nearbyint implementations wrongly clear
2306 all exceptions
2307 [15786] dynamic-link: ifunc resolver functions can smash function
2308 arguments
2309 [15918] math: Unnecessary check for equality in hypotf()
2310 [16061] localedata: Review / update transliteration data
2311 [16068] math: [i386/x86_64] x86 and x86_64 fesetenv exclude state they
2312 should include
2313 [16141] time: strptime %z offset restriction
2314 [16171] math: drem should be alias of remainder
2315 [16296] math: fegetround is pure?
2316 [16347] math: [ldbl-128ibm] ldbl-128/e_lgammal_r.c may not be suitable.
2317 [16364] libc: sleep may leave SIGCHLD blocked on sync cancellation on
2318 GNU/Linux
2319 [16399] math: [mips] lrint / llrint / lround / llround missing exceptions
2320 [16415] math: Clean up ldbl-128 / ldbl-128ibm expm1l for large positive
2321 arguments
2322 [16422] math: [powerpc] math-float, math-double failing llrint tests with
2323 "Exception "Inexact" set" on ppc32
2324 [16495] localedata: nl_NL: date_fmt: shuffle year/month around
2325 [16517] math: Missing underflow exception from tanf/tan/tanl
2326 [16519] math: Missing underflow exception from sinhf
2327 [16520] math: Missing underflow exception from tanhf
2328 [16521] math: Missing underflow exception from exp2
2329 [16620] math: [ldbl-128ibm] exp10l spurious overflows / bad directed
2330 rounding results
2331 [16734] stdio: fopen calls mmap to allocate its buffer
2332 [16961] math: nan function incorrect handling of bad sequences
2333 [16962] math: nan function unbounded stack allocation (CVE-2014-9761)
2334 [16973] localedata: Fix lang_lib/lang_term as per ISO 639-2
2335 [16985] locale: localedef: confusing error message when opening output
2336 fails
2337 [17118] math: ctanh(INFINITY + 2 * I) returns incorrect value
2338 [17197] locale: Redundant shift character in iconv conversion output at
2339 block boundary
2340 [17243] libc: trunk/posix/execl.c:53: va_args problem ?
2341 [17244] libc: trunk/sysdeps/unix/sysv/linux/semctl.c:116: va_args muxup ?
2342 [17250] dynamic-link: static linking breaks nss loading
2343 (getaddrinfo/getpwnam/etc...)
2344 [17404] libc: atomic_exchange_rel lacking a barrier on MIPS16, GCC before
2345 4.7?
2346 [17441] math: isnan() should use __builtin_isnan() in GCC
2347 [17514] nptl: Assert failure unlocking ERRORCHECK mutex after timedlock
2348 (related to lock elision)
2349 [17787] manual: Exponent on page 324 of the PDF ends prematurely
2350 [17886] time: strptime should be able to parse "Z" as a timezone with %z
2351 [17887] time: strptime should be able to parse "+01:00" style timezones
2352 [17905] libc: catopen() Multiple unbounded stack allocations
2353 (CVE-2015-8779)
2354 [18084] libc: backtrace (..., 0) dumps core on x86
2355 [18086] libc: nice() sets errno to 0 on success
2356 [18240] libc: hcreate, hcreate_r should fail with ENOMEM if element count
2357 is too large (CVE-2015-8778)
2358 [18251] dynamic-link: SONAME missing when audit modules provides path
2359 [18265] libc: add attributes for wchar string and memory functions
2360 [18370] math: csqrt missing underflows
2361 [18421] libc: [hppa] read-only segment has dynamic relocations
2362 [18472] libc: Obsolete syscall wrappers should be compat symbols
2363 [18480] libc: hppa glibc miscompilation in sched_setaffinity()
2364 [18491] localedata: Update tr_TR LC_CTYPE as part of Unicode updates
2365 [18525] localedata: Remove locale timezone information
2366 [18560] libc: [powerpc] spurious bits/ipc.h definitions
2367 [18568] localedata: Update locale data to Unicode 8.0
2368 [18589] locale: sort-test.sh fails at random
2369 [18595] math: ctan, ctanh missing underflows
2370 [18604] libc: assert macro-expands its argument
2371 [18610] math: S390: fetestexcept() reports any exception if DXC-code
2372 contains a vector instruction exception.
2373 [18611] math: j1, jn missing errno setting on underflow
2374 [18618] localedata: sync Chechen locale definitions with other *_RU
2375 locales
2376 [18647] math: powf(-0x1.000002p0, 0x1p30) returns 0 instead of +inf
2377 [18661] libc: Some x86-64 assembly codes don't align stack to 16 bytes
2378 [18665] network: In send_dg, the recvfrom function is NOT always using the
2379 buffer size of a newly created buffer (CVE-2015-7547)
2380 [18674] libc: [i386] trunk/sysdeps/i386/tst-auditmod3b.c:84: possible
2381 missing break ?
2382 [18675] libc: fpathconf(_PC_NAME_MAX) fails against large filesystems for
2383 32bit processes
2384 [18681] libc: regexp.h is obsolete and buggy, and should be desupported
2385 [18699] math: tilegx cproj() for various complex infinities does not yield
2386 infinity
2387 [18724] libc: Harden put*ent functions against data injection
2388 [18743] nptl: PowerPC: findutils testcase fails with --enable-lock-elision
2389 [18755] build: build errors with -DNDEBUG
2390 [18757] stdio: fmemopen fails to set errno on failure
2391 [18778] dynamic-link: ld.so crashes if failed dlopen causes libpthread to
2392 be forced unloaded
2393 [18781] libc: openat64 lacks O_LARGEFILE
2394 [18787] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/atomic.h:71:6:
2395 error: can’t find a register in class ‘R1_REGS’ while reloading ‘asm’
2396 [18789] math: [ldbl-128ibm] sinhl inaccurate near 0
2397 [18790] math: [ldbl-128ibm] tanhl inaccurate
2398 [18795] libc: stpncpy fortification misses buffer lengths that are
2399 statically too large
2400 [18796] build: build fails for --disable-mathvec
2401 [18803] math: hypot missing underflows
2402 [18820] stdio: fmemopen may leak memory on failure
2403 [18823] math: csqrt spurious underflows
2404 [18824] math: fma spurious underflows
2405 [18825] math: pow missing underflows
2406 [18857] math: [ldbl-128ibm] nearbyintl wrongly uses signaling comparisons
2407 [18868] nptl: pthread_barrier_init typo has in-theory-undefined behavior
2408 [18870] build: sem_open.c fails to compile with missing symbol
2409 FUTEX_SHARED
2410 [18872] stdio: Fix memory leak in printf_positional
2411 [18873] libc: posix_fallocate overflow check ineffective
2412 [18875] math: Excess precision leads incorrect libm
2413 [18877] libc: arm: mmap offset regression
2414 [18887] libc: memory corruption when using getmntent on blank lines
2415 [18918] localedata: hu_HU: change time to HH:MM:SS format
2416 [18921] libc: Regression: extraneous stat() and fstat() performed by
2417 opendir()
2418 [18928] dynamic-link: LD_POINTER_GUARD is not ignored for privileged
2419 binaries (CVE-2015-8777)
2420 [18951] math: tgamma missing underflows
2421 [18952] math: [ldbl-128/ldbl-128ibm] lgammal spurious "invalid", incorrect
2422 signgam
2423 [18953] localedata: lt_LT: change currency symbol to the euro
2424 [18956] math: powf inaccuracy
2425 [18961] math: [i386] exp missing underflows
2426 [18966] math: [i386] exp10 missing underflows
2427 [18967] math: math.h XSI POSIX namespace (gamma, isnan, scalb)
2428 [18969] build: multiple string test failures due to missing locale
2429 dependencies
2430 [18970] libc: Reference of pthread_setcancelstate in libc.a
2431 [18977] math: float / long double Bessel functions not in XSI POSIX
2432 [18980] math: i386 libm functions return with excess range and precision
2433 [18981] math: i386 scalb*, ldexp return with excess range and precision
2434 [18982] stdio: va_list and vprintf
2435 [18985] time: Passing out of range data to strftime() causes a segfault
2436 (CVE-2015-8776)
2437 [19003] math: [x86_64] fma4 version of pow inappropriate contraction
2438 [19007] libc: FAIL: elf/check-localplt with -z now and binutils 2.26
2439 [19012] locale: iconv_open leaks memory on error path
2440 [19016] math: clog, clog10 inaccuracy
2441 [19018] nptl: Mangle function pointers in tls_dtor_list
2442 [19032] math: [i386] acosh (-qNaN) spurious "invalid" exception
2443 [19046] math: ldbl-128 / ldbl-128ibm lgamma bad overflow handling
2444 [19048] malloc: malloc: arena free list can become cyclic, increasing
2445 contention
2446 [19049] math: [powerpc] erfc incorrect zero sign
2447 [19050] math: [powerpc] log* incorrect zero sign
2448 [19058] math: [x86_64] Link fail with -fopenmp and -flto
2449 [19059] math: nexttoward overflow incorrect in non-default rounding modes
2450 [19071] math: ldbl-96 lroundl incorrect just below powers of 2
2451 [19074] network: Data race in _res_hconf_reorder_addrs
2452 [19076] math: [ldbl-128ibm] log1pl (-1) wrong sign of infinity
2453 [19077] math: [ldbl-128ibm] logl (1) incorrect sign of zero result
2454 [19078] math: [ldbl-128ibm] expl overflow incorrect in non-default
2455 rounding modes
2456 [19079] math: dbl-64/wordsize-64 lround based on llround incorrect for
2457 ILP32
2458 [19085] math: ldbl-128 lrintl, lroundl missing exceptions for 32-bit long
2459 [19086] manual: posix_fallocate64 documented argument order is wrong.
2460 [19088] math: lround, llround missing exceptions close to overflow
2461 threshold
2462 [19094] math: lrint, llrint missing exceptions close to overflow threshold
2463 [19095] math: dbl-64 lrint incorrect for 64-bit long
2464 [19122] dynamic-link: Unnecessary PLT relocations in librtld.os
2465 [19124] dynamic-link: ld.so failed to build with older assmebler
2466 [19125] math: [powerpc32] llroundf, llround incorrect exceptions
2467 [19129] dynamic-link: [arm] Concurrent lazy TLSDESC resolution can crash
2468 [19134] math: [powerpc32] lround, lroundf spurious exceptions
2469 [19137] libc: i386/epoll_pwait.S doesn't support cancellation
2470 [19143] nptl: Remove CPU set size checking from sched_setaffinity,
2471 pthread_setaffinity_np
2472 [19156] math: [ldbl-128] j0l spurious underflows
2473 [19164] nptl: tst-getcpu fails with many possible CPUs
2474 [19168] math: math/test-ildoubl and math/test-ldouble failure
2475 [19174] nptl: PowerPC: TLE enabled pthread mutex performs poorly.
2476 [19178] dynamic-link: ELF_RTYPE_CLASS_EXTERN_PROTECTED_DATA confuses
2477 prelink
2478 [19181] math: [i386/x86_64] fesetenv (FE_DFL_ENV), fesetenv
2479 (FE_NOMASK_ENV) do not clear SSE exceptions
2480 [19182] malloc: malloc deadlock between ptmalloc_lock_all and
2481 _int_new_arena/reused_arena
2482 [19189] math: [ldbl-128] log1pl (-qNaN) spurious "invalid" exception
2483 [19201] math: dbl-64 remainder incorrect sign of zero result
2484 [19205] math: bits/math-finite.h conditions do not match math.h and
2485 bits/mathcalls.h
2486 [19209] math: bits/math-finite.h wrongly maps ldexp to scalbn
2487 [19211] math: lgamma functions do not set signgam for -ffinite-math-only
2488 for C99-based standards
2489 [19212] libc: features.h not -Wundef clean
2490 [19213] math: [i386/x86_64] log* (1) incorrect zero sign for -ffinite-
2491 math-only
2492 [19214] libc: Family and model identification for AMD CPU's are incorrect.
2493 [19219] libc: GLIBC build fails for ia64 with missing __nearbyintl
2494 [19228] math: [powerpc] nearbyint wrongly clears "inexact", leaves traps
2495 disabled
2496 [19235] math: [powerpc64] lround, lroundf, llround, llroundf spurious
2497 "inexact" exceptions
2498 [19238] math: [powerpc] round, roundf spurious "inexact" for integer
2499 arguments
2500 [19242] libc: strtol incorrect in Turkish locales
2501 [19243] malloc: reused_arena can pick an arena on the free list, leading
2502 to an assertion failure and reference count corruption
2503 [19253] time: tzset() ineffective when temporary TZ did not include DST
2504 rules
2505 [19266] math: strtod ("NAN(I)") incorrect in Turkish locales
2506 [19270] math: [hppa] Shared libm missing __isnanl
2507 [19285] libc: [hppa] sysdeps/unix/sysv/linux/hppa/bits/mman.h: missing
2508 MAP_HUGETLB and MAP_STACK defines
2509 [19313] nptl: Wrong __cpu_mask for x32
2510 [19347] libc: grantpt: try to force a specific gid even without pt_chown
2511 [19349] math: [ldbl-128ibm] tanhl inaccurate for small arguments
2512 [19350] math: [ldbl-128ibm] sinhl spurious overflows
2513 [19351] math: [ldbl-128ibm] logl inaccurate near 1
2514 [19363] time: x32: times() return value wrongly truncates/sign extends
2515 from 32bit
2516 [19367] dynamic-link: Improve branch prediction on Silvermont
2517 [19369] network: Default domain name not reset by res_ninit when "search"
2518 / "domain" entry is removed from resolv.conf
2519 [19375] math: powerpc: incorrect results for POWER7 logb with negative
2520 subnormals
2521 [19385] localedata: bg_BG: time separator should be colon, not comma
2522 [19408] libc: linux personality syscall wrapper may erroneously return an
2523 error on 32-bit architectures
2524 [19415] libc: dladdr returns wrong names on hppa
2525 [19432] libc: iconv rejects redundant escape sequences in IBM900, IBM903,
2526 IBM905, IBM907, and IBM909
2527 [19439] math: Unix98 isinf and isnan functions conflict with C++11
2528 [19443] build: build failures with -DDEBUG
2529 [19451] build: Make check fails on test-double-vlen2
2530 [19462] libc: Glibc failed to build with -Os
2531 [19465] math: Wrong code with -Os
2532 [19466] time: time/tst-mktime2.c is compiled into an infinite loop with
2533 -Os
2534 [19467] string: Fast_Unaligned_Load needs to be enabled for Excavator core
2535 CPU's.
2536 [19475] libc: Glibc 2.22 doesn't build on sparc [PATCH]
2537 [19486] math: S390: Math tests fail with "Exception Inexact set".
2538 [19529] libc: [ARM]: FAIL: stdlib/tst-makecontext
2539 [19550] libc: [mips] mmap negative offset handling inconsistent with other
2540 architectures
2541 [19590] math: Fail to build shared objects that use libmvec.so functions.
2542 \f
2543 Version 2.22
2544
2545 * The following bugs are resolved with this release:
2546
2547 438, 2981, 4719, 6544, 6792, 11216, 12836, 13028, 13064, 13151, 13152,
2548 14094, 14113, 14292, 14841, 14906, 14958, 15319, 15467, 15790, 15969,
2549 16159, 16339, 16350, 16351, 16352, 16353, 16361, 16512, 16526, 16538,
2550 16559, 16560, 16704, 16783, 16850, 17053, 17090, 17195, 17269, 17293,
2551 17322, 17403, 17475, 17523, 17542, 17569, 17581, 17588, 17596, 17620,
2552 17621, 17628, 17631, 17692, 17711, 17715, 17776, 17779, 17792, 17833,
2553 17836, 17841, 17912, 17916, 17930, 17932, 17944, 17949, 17964, 17965,
2554 17967, 17969, 17977, 17978, 17987, 17991, 17996, 17998, 17999, 18007,
2555 18019, 18020, 18029, 18030, 18032, 18034, 18036, 18038, 18039, 18042,
2556 18043, 18046, 18047, 18049, 18068, 18078, 18080, 18093, 18100, 18104,
2557 18110, 18111, 18116, 18125, 18128, 18134, 18138, 18185, 18196, 18197,
2558 18206, 18210, 18211, 18217, 18219, 18220, 18221, 18234, 18244, 18245,
2559 18247, 18287, 18319, 18324, 18333, 18346, 18371, 18383, 18397, 18400,
2560 18409, 18410, 18412, 18418, 18422, 18434, 18444, 18457, 18468, 18469,
2561 18470, 18479, 18483, 18495, 18496, 18497, 18498, 18502, 18507, 18508,
2562 18512, 18513, 18519, 18520, 18522, 18527, 18528, 18529, 18530, 18532,
2563 18533, 18534, 18536, 18539, 18540, 18542, 18544, 18545, 18546, 18547,
2564 18549, 18553, 18557, 18558, 18569, 18583, 18585, 18586, 18592, 18593,
2565 18594, 18602, 18612, 18613, 18619, 18633, 18635, 18641, 18643, 18648,
2566 18657, 18676, 18694, 18696, 18887.
2567
2568 * Cache information can be queried via sysconf() function on s390 e.g. with
2569 _SC_LEVEL1_ICACHE_SIZE as argument.
2570
2571 * A buffer overflow in gethostbyname_r and related functions performing DNS
2572 requests has been fixed. If the NSS functions were called with a
2573 misaligned buffer, the buffer length change due to pointer alignment was
2574 not taken into account. This could result in application crashes or,
2575 potentially arbitrary code execution, using crafted, but syntactically
2576 valid DNS responses. (CVE-2015-1781)
2577
2578 * The time zone file parser has been made more robust against crafted time
2579 zone files, avoiding heap buffer overflows related to the processing of
2580 the tzh_ttisstdcnt and tzh_ttisgmtcnt fields, and a stack overflow due to
2581 large time zone data files. Overly long time zone specifiers in the TZ
2582 variable no longer result in stack overflows and crashes.
2583
2584 * A powerpc and powerpc64 optimization for TLS, similar to TLS descriptors
2585 for LD and GD on x86 and x86-64, has been implemented. You will need
2586 binutils-2.24 or later to enable this optimization.
2587
2588 * Character encoding and ctype tables were updated to Unicode 7.0.0, using
2589 new generator scripts contributed by Pravin Satpute and Mike FABIAN (Red
2590 Hat). These updates cause user visible changes, such as the fix for bug
2591 17998.
2592
2593 * CVE-2014-8121 The NSS backends shared internal state between the getXXent
2594 and getXXbyYY NSS calls for the same database, causing a denial-of-service
2595 condition in some applications.
2596
2597 * Added vector math library named libmvec with the following vectorized x86_64
2598 implementations: cos, cosf, sin, sinf, sincos, sincosf, log, logf, exp, expf,
2599 pow, powf.
2600 The library can be disabled with --disable-mathvec. Use of the functions is
2601 enabled with -fopenmp -ffast-math starting from -O1 for GCC version >= 4.9.0.
2602 Shared library libmvec.so is linked in as needed when using -lm (no need to
2603 specify -lmvec explicitly for not static builds).
2604 Visit <https://sourceware.org/glibc/wiki/libmvec> for detailed information.
2605
2606 * A new fmemopen implementation has been added with the goal of POSIX
2607 compliance. The new implementation fixes the following long-standing
2608 issues: BZ#6544, BZ#11216, BZ#12836, BZ#13151, BZ#13152, and BZ#14292. The
2609 old implementation is still present for use be by existing binaries.
2610
2611 * The 32-bit sparc sigaction ABI was inadvertently broken in the 2.20 and 2.21
2612 releases. It has been fixed to match 2.19 and older, but binaries built
2613 against 2.20 and 2.21 might need to be recompiled. See BZ#18694.
2614
2615 * Port to Native Client running on ARMv7-A (--host=arm-nacl).
2616 Contributed by Roland McGrath (Google).
2617
2618 * The header <regexp.h> is deprecated, and will be removed in a future
2619 release. Use of this header will trigger a deprecation warning.
2620 Application developers should update their code to use <regex.h> instead.
2621
2622 This header was formerly part of SUS, but was deprecated in 1994 and
2623 removed from the standard in 2001. Also, the glibc implementation
2624 leaks memory. See BZ#18681 for more details.
2625 \f
2626 Version 2.21
2627
2628 * The following bugs are resolved with this release:
2629
2630 6652, 10672, 12674, 12847, 12926, 13862, 14132, 14138, 14171, 14498,
2631 15215, 15378, 15884, 16009, 16418, 16191, 16469, 16576, 16617, 16618,
2632 16619, 16657, 16740, 16857, 17192, 17266, 17273, 17344, 17363, 17370,
2633 17371, 17411, 17460, 17475, 17485, 17501, 17506, 17508, 17522, 17555,
2634 17570, 17571, 17572, 17573, 17574, 17582, 17583, 17584, 17585, 17589,
2635 17594, 17601, 17608, 17616, 17625, 17630, 17633, 17634, 17635, 17647,
2636 17653, 17657, 17658, 17664, 17665, 17668, 17682, 17702, 17717, 17719,
2637 17722, 17723, 17724, 17725, 17732, 17733, 17744, 17745, 17746, 17747,
2638 17748, 17775, 17777, 17780, 17781, 17782, 17791, 17793, 17796, 17797,
2639 17801, 17803, 17806, 17834, 17844, 17848, 17868, 17869, 17870, 17885,
2640 17892.
2641
2642 * CVE-2015-1472 CVE-2015-1473 Under certain conditions wscanf can allocate
2643 too little memory for the to-be-scanned arguments and overflow the
2644 allocated buffer. The implementation now correctly computes the required
2645 buffer size when using malloc, and switches to malloc from alloca as
2646 intended.
2647
2648 * A new semaphore algorithm has been implemented in generic C code for all
2649 machines. Previous custom assembly implementations of semaphore were
2650 difficult to reason about or ensure that they were safe. The new version
2651 of semaphore supports machines with 64-bit or 32-bit atomic operations.
2652 The new semaphore algorithm is used by sem_init, sem_open, sem_post,
2653 sem_wait, sem_timedwait, sem_trywait, and sem_getvalue.
2654
2655 * Port to Altera Nios II has been contributed by Mentor Graphics.
2656
2657 * Optimized strcpy, stpcpy, strncpy, stpncpy, strcmp, and strncmp
2658 implementations for powerpc64/powerpc64le.
2659 Implemented by Adhemerval Zanella (IBM).
2660
2661 * Added support for TSX lock elision of pthread mutexes on powerpc32, powerpc64
2662 and powerpc64le. This may improve lock scaling of existing programs on
2663 HTM capable systems. The lock elision code is only enabled with
2664 --enable-lock-elision=yes. Also, the TSX lock elision implementation for
2665 powerpc will issue a transaction abort on every syscall to avoid side
2666 effects being visible outside transactions.
2667
2668 * Optimized strcpy, stpcpy, strchrnul and strrchr implementations for
2669 AArch64. Contributed by ARM Ltd.
2670
2671 * i386 memcpy functions optimized with SSE2 unaligned load/store.
2672
2673 * CVE-2104-7817 The wordexp function could ignore the WRDE_NOCMD flag
2674 under certain input conditions resulting in the execution of a shell for
2675 command substitution when the application did not request it. The
2676 implementation now checks WRDE_NOCMD immediately before executing the
2677 shell and returns the error WRDE_CMDSUB as expected.
2678
2679 * CVE-2012-3406 printf-style functions could run into a stack overflow when
2680 processing format strings with a large number of format specifiers.
2681
2682 * CVE-2014-9402 The nss_dns implementation of getnetbyname could run into an
2683 infinite loop if the DNS response contained a PTR record of an unexpected
2684 format.
2685
2686 * The minimum GCC version that can be used to build this version of the GNU
2687 C Library is GCC 4.6. Older GCC versions, and non-GNU compilers, can
2688 still be used to compile programs using the GNU C Library.
2689
2690 * The GNU C Library is now built with -Werror by default. This can be
2691 disabled by configuring with --disable-werror.
2692
2693 * New locales: tu_IN, bh_IN, raj_IN, ce_RU.
2694
2695 * The obsolete sigvec function has been removed. This was the original
2696 4.2BSD interface that inspired the POSIX.1 sigaction interface, which
2697 programs have been using instead for about 25 years. Of course, ABI
2698 compatibility for old binaries using sigvec remains intact.
2699
2700 * Merged gettext 0.19.3 into the intl subdirectory. This fixes building
2701 with newer versions of bison.
2702
2703 * Support for MIPS o32 FPXX, FP64A and FP64 ABI Extensions.
2704 The original MIPS o32 hard-float ABI requires an FPU where double-precision
2705 registers overlay two consecutive single-precision registers. MIPS32R2
2706 introduced a new FPU mode (FR=1) where double-precision registers extend the
2707 corresponding single-precision registers which is incompatible with the
2708 o32 hard-float ABI. The MIPS SIMD ASE and the MIPSR6 architecture both
2709 require the use of FR=1 making a transition necessary. New o32 ABI
2710 extensions enable users to migrate over time from the original o32 ABI
2711 through to the updated o32 FP64 ABI. To achieve this the dynamic linker now
2712 tracks the ABI of any loaded object and verifies that new objects are
2713 compatible. Mode transitions will also be requested as required and
2714 unsupportable objects will be rejected. The ABI checks include both soft and
2715 hard float ABIs for o32, n32 and n64.
2716
2717 GCC 5 with GNU binutils 2.25 onwards:
2718 It is strongly recommended that all o32 system libraries are built using the
2719 new o32 FPXX ABI (-mfpxx) to facilitate the transition as this is compatible
2720 with the original and all new o32 ABI extensions. Configure a MIPS GCC
2721 compiler using --with-fp-32=xx to set this by default.
2722 \f
2723 Version 2.20
2724
2725 * The following bugs are resolved with this release:
2726
2727 6804, 9894, 11505, 12994, 13347, 13651, 14308, 14770, 15119, 15132, 15347,
2728 15514, 15698, 15804, 15894, 15946, 16002, 16064, 16095, 16194, 16198,
2729 16275, 16284, 16287, 16315, 16348, 16349, 16354, 16357, 16362, 16447,
2730 16516, 16532, 16539, 16545, 16561, 16562, 16564, 16574, 16599, 16600,
2731 16609, 16610, 16611, 16613, 16619, 16623, 16629, 16632, 16634, 16639,
2732 16642, 16648, 16649, 16670, 16674, 16677, 16680, 16681, 16683, 16689,
2733 16695, 16701, 16706, 16707, 16712, 16713, 16714, 16724, 16731, 16739,
2734 16740, 16743, 16754, 16758, 16759, 16760, 16770, 16786, 16789, 16791,
2735 16796, 16799, 16800, 16815, 16823, 16824, 16831, 16838, 16839, 16849,
2736 16854, 16876, 16877, 16878, 16882, 16885, 16888, 16890, 16892, 16912,
2737 16915, 16916, 16917, 16918, 16922, 16927, 16928, 16932, 16943, 16958,
2738 16965, 16966, 16967, 16977, 16978, 16984, 16990, 16996, 17009, 17022,
2739 17031, 17042, 17048, 17050, 17058, 17061, 17062, 17069, 17075, 17078,
2740 17079, 17084, 17086, 17088, 17092, 17097, 17125, 17135, 17137, 17150,
2741 17153, 17187, 17213, 17259, 17261, 17262, 17263, 17319, 17325, 17354.
2742
2743 * Reverted change of ABI data structures for s390 and s390x:
2744 On s390 and s390x the size of struct ucontext and jmp_buf was increased in
2745 2.19. This change is reverted in 2.20. The introduced 2.19 symbol versions
2746 of getcontext, setjmp, _setjmp, __sigsetjmp, longjmp, _longjmp, siglongjmp
2747 are preserved pointing straight to the same implementation as the old ones.
2748 Given that, new callers will simply provide a too-big buffer to these
2749 functions. Any applications/libraries out there that embed jmp_buf or
2750 ucontext_t in an ABI-relevant data structure that have already been rebuilt
2751 against 2.19 headers will have to rebuilt again. This is necessary in any
2752 case to revert the breakage in their ABI caused by the glibc change.
2753
2754 * Support for file description locks is added to systems running the
2755 Linux kernel. The standard file locking interfaces are extended to
2756 operate on file descriptions, not file descriptors, via the use of
2757 F_OFD_GETLK, F_OFD_SETLK, and F_OFD_SETLKW. File description locks
2758 are associated with an open file instead of a process.
2759
2760 * Optimized strchr implementation for AArch64. Contributed by ARM Ltd.
2761
2762 * The minimum Linux kernel version that this version of the GNU C Library
2763 can be used with is 2.6.32.
2764
2765 * Running the testsuite no longer terminates as soon as a test fails.
2766 Instead, a file tests.sum (xtests.sum from "make xcheck") is generated,
2767 with PASS or FAIL lines for individual tests. A summary of the results is
2768 printed, including a list of failing lists, and "make check" exits with
2769 error status if there were any unexpected failures. "make check
2770 stop-on-test-failure=y" may be used to keep the old behavior.
2771
2772 * The am33 port, which had not worked for several years, has been removed
2773 from ports.
2774
2775 * The _BSD_SOURCE and _SVID_SOURCE feature test macros are no longer
2776 supported; they now act the same as _DEFAULT_SOURCE (but generate a
2777 warning). Except for cases where _BSD_SOURCE enabled BSD interfaces that
2778 conflicted with POSIX (support for which was removed in 2.19), the
2779 interfaces those macros enabled remain available when compiling with
2780 _GNU_SOURCE defined, with _DEFAULT_SOURCE defined, or without any feature
2781 test macros defined.
2782
2783 * Optimized strcmp implementation for ARMv7. Contributed by ARM Ltd.
2784
2785 * Added support for TX lock elision of pthread mutexes on s390 and s390x.
2786 This may improve lock scaling of existing programs on TX capable systems.
2787 The lock elision code is only built with --enable-lock-elision=yes and
2788 then requires a GCC version supporting the TX builtins. With lock elision
2789 default mutexes are elided via __builtin_tbegin, if the cpu supports
2790 transactions. By default lock elision is not enabled and the elision code
2791 is not built.
2792
2793 * CVE-2014-4043 The posix_spawn_file_actions_addopen implementation did not
2794 copy the path argument. This allowed programs to cause posix_spawn to
2795 deference a dangling pointer, or use an unexpected pathname argument if
2796 the string was modified after the posix_spawn_file_actions_addopen
2797 invocation.
2798
2799 * All supported architectures now use the main glibc sysdeps directory
2800 instead of some being in a separate "ports" directory (which was
2801 distributed separately before glibc 2.17).
2802
2803 * The NPTL implementation of POSIX pthreads is no longer an "add-on".
2804 On configurations that support it (all Linux configurations), it's now
2805 used regardless of the --enable-add-ons switch to configure. It is no
2806 longer possible to build such configurations without pthreads support.
2807
2808 * Locale names, including those obtained from environment variables (LANG
2809 and the LC_* variables), are more tightly checked for proper syntax.
2810 setlocale will now fail (with EINVAL) for locale names that are overly
2811 long, contain slashes without starting with a slash, or contain ".." path
2812 components. (CVE-2014-0475) Previously, some valid locale names were
2813 silently replaced with the "C" locale when running in AT_SECURE mode
2814 (e.g., in a SUID program). This is no longer necessary because of the
2815 additional checks.
2816
2817 * On x86-64, the dynamic linker's lazy-binding support is now compatible
2818 with application code using Intel MPX instructions. (With all previous
2819 versions, the MPX register state could be clobbered when making calls
2820 into or out of a shared library.) Note that while the new dynamic
2821 linker is compatible with all known x86 hardware whether or not it
2822 supports Intel MPX, some x86 instruction-set emulators might fail to
2823 handle the new instruction encodings. This is known to affect Valgrind
2824 versions up through 3.9 (but will be fixed in the forthcoming 3.10
2825 release), and might affect other tools that do instruction emulation.
2826
2827 * Support for loadable gconv transliteration modules has been removed.
2828 The support for transliteration modules has been non-functional for
2829 over a decade, and the removal is prompted by security defects. The
2830 normal gconv conversion modules are still supported. Transliteration
2831 with //TRANSLIT is still possible, and the //IGNORE specifier
2832 continues to be supported. (CVE-2014-5119)
2833
2834 * Decoding a crafted input sequence in the character sets IBM933, IBM935,
2835 IBM937, IBM939, IBM1364 could result in an out-of-bounds array read,
2836 resulting a denial-of-service security vulnerability in applications which
2837 use functions related to iconv. (CVE-2014-6040)
2838 \f
2839 Version 2.19
2840
2841 * The following bugs are resolved with this release:
2842
2843 156, 387, 431, 762, 832, 926, 2801, 4772, 6786, 6787, 6807, 6810, 6981,
2844 7003, 9721, 9954, 10253, 10278, 11087, 11157, 11214, 12100, 12486, 12751,
2845 12986, 13028, 13982, 13985, 14029, 14032, 14120, 14143, 14155, 14286,
2846 14547, 14699, 14752, 14782, 14876, 14910, 15004, 15048, 15073, 15089,
2847 15128, 15218, 15268, 15277, 15308, 15362, 15374, 15400, 15425, 15427,
2848 15483, 15522, 15531, 15532, 15593, 15601, 15608, 15609, 15610, 15632,
2849 15640, 15670, 15672, 15680, 15681, 15723, 15734, 15735, 15736, 15748,
2850 15749, 15754, 15760, 15763, 15764, 15797, 15799, 15825, 15843, 15844,
2851 15846, 15847, 15849, 15850, 15855, 15856, 15857, 15859, 15867, 15886,
2852 15887, 15890, 15892, 15893, 15895, 15897, 15901, 15905, 15909, 15915,
2853 15917, 15919, 15921, 15923, 15939, 15941, 15948, 15963, 15966, 15968,
2854 15985, 15988, 15997, 16032, 16034, 16036, 16037, 16038, 16041, 16046,
2855 16055, 16071, 16072, 16074, 16077, 16078, 16103, 16112, 16143, 16144,
2856 16146, 16150, 16151, 16153, 16167, 16169, 16172, 16195, 16214, 16245,
2857 16271, 16274, 16283, 16289, 16293, 16314, 16316, 16330, 16337, 16338,
2858 16356, 16365, 16366, 16369, 16372, 16375, 16379, 16381, 16384, 16385,
2859 16386, 16387, 16390, 16394, 16398, 16400, 16407, 16408, 16414, 16430,
2860 16431, 16453, 16474, 16506, 16510, 16529
2861
2862 * Slovenian translations for glibc messages have been contributed by the
2863 Translation Project's Slovenian team of translators.
2864
2865 * The public headers no longer use __unused nor __block. This change is to
2866 support compiling programs that are derived from BSD sources and use
2867 __unused internally, and to support compiling with Clang's -fblock
2868 extension which uses __block.
2869
2870 * CVE-2012-4412 The strcoll implementation caches indices and rules for
2871 large collation sequences to optimize multiple passes. This cache
2872 computation may overflow for large collation sequences and may cause a
2873 stack or buffer overflow. This is now fixed to use a slower algorithm
2874 which does not use a cache if there is an integer overflow.
2875
2876 * CVE-2012-4424 The strcoll implementation uses malloc to cache indices and
2877 rules for large collation sequences to optimize multiple passes and falls
2878 back to alloca if malloc fails, resulting in a possible stack overflow.
2879 The implementation now falls back to an uncached collation sequence lookup
2880 if malloc fails.
2881
2882 * CVE-2013-4788 The pointer guard used for pointer mangling was not
2883 initialized for static applications resulting in the security feature
2884 being disabled. The pointer guard is now correctly initialized to a
2885 random value for static applications. Existing static applications need
2886 to be recompiled to take advantage of the fix (bug 15754).
2887
2888 * CVE-2013-4237 The readdir_r function could write more than NAME_MAX bytes
2889 to the d_name member of struct dirent, or omit the terminating NUL
2890 character. (Bugzilla #14699).
2891
2892 * CVE-2013-4332 The pvalloc, valloc, memalign, posix_memalign and
2893 aligned_alloc functions could allocate too few bytes or corrupt the
2894 heap when passed very large allocation size values (Bugzilla #15855,
2895 #15856, #15857).
2896
2897 * CVE-2013-4458 Stack overflow in getaddrinfo with large number of results
2898 for AF_INET6 has been fixed (Bugzilla #16072).
2899
2900 * New locales: ak_GH, anp_IN, ar_SS, cmn_TW, hak_TW, lzh_TW, nan_TW, pap_AW,
2901 pap_CW, quz_PE, the_NP.
2902
2903 * Substantially revised locales: gd_GB, ht_HT
2904
2905 * The LC_ADDRESS field was updated to support country_car for almost all
2906 supported locales.
2907
2908 * ISO 1427 definitions were updated.
2909
2910 * ISO 3166 definitions were updated.
2911
2912 * The localedef utility now supports --big-endian and --little-endian
2913 command-line options to generate locales for a different system from that
2914 for which the C library was built.
2915
2916 * Binary locale files now only depend on the endianness of the system for
2917 which they are generated and not on other properties of that system. As a
2918 consequence, binary files generated with new localedef may be incompatible
2919 with old versions of the GNU C Library, and binary files generated with
2920 old localedef may be incompatible with this version of the GNU C Library,
2921 in the following circumstances:
2922
2923 + Locale files may be incompatible on m68k systems.
2924
2925 + Locale archive files (but not separate files for individual locales) may
2926 be incompatible on systems where plain "char" is signed.
2927
2928 * The configure option --disable-versioning has been removed. Builds with
2929 --disable-versioning had not worked for several years.
2930
2931 * ISO 639 definitions were updated for Chiga (cgg) and Chinese (gan, hak, czh,
2932 cjy, lzh, cmn, mnp, cdo, czo, cpx, wuu, hsn, yue).
2933
2934 * SystemTap probes for malloc have been introduced.
2935
2936 * SystemTap probes for slow multiple precision fallback paths of
2937 transcendental functions have been introduced.
2938
2939 * Support for powerpc64le has been added.
2940
2941 * The soft-float powerpc port now supports e500 processors.
2942
2943 * Support for STT_GNU_IFUNC symbols added for ppc32/power4+ and ppc64.
2944
2945 * A new feature test macro _DEFAULT_SOURCE is available to enable the same
2946 set of header declarations that are enabled by default, even when other
2947 feature test macros or compiler options such as -std=c99 would otherwise
2948 disable some of those declarations.
2949
2950 * The _BSD_SOURCE feature test macro no longer enables BSD interfaces that
2951 conflict with POSIX. The libbsd-compat library (which was a dummy library
2952 that did nothing) has also been removed.
2953
2954 * Preliminary documentation about Multi-Thread, Async-Signal and
2955 Async-Cancel Safety has been added.
2956
2957 * Change of ABI data structures for s390 and s390x:
2958 On s390 and s390x the size of struct ucontext and jmp_buf was increased to
2959 allow for future hardware extensions. All library functions that accept or
2960 return these structures were versioned in order to provide backwards
2961 compatibility with existing code. However, not all cases can be handled
2962 with interface versioning. If an application embeds either structure into
2963 another structure and that structure is passed to another compilation unit
2964 which expects the newer larger-sized object then any access to the new
2965 fields will result in undefined behaviour. Similarly any access to fields
2966 that were moved by the enlarging of the structures will cause undefined
2967 behaviour. To fix the undefined behaviour all objects that are part of the
2968 final in-memory image must agree on the size of structures, and this may
2969 require recompilation.
2970 \f
2971 Version 2.18
2972
2973 * The following bugs are resolved with this release:
2974
2975 2546, 2560, 5159, 6809, 7006, 10060, 10062, 10283, 10357, 10686, 11120,
2976 11561, 12310, 12387, 12492, 12515, 12723, 13550, 13889, 13951, 13988,
2977 14142, 14176, 14200, 14256, 14280, 14293, 14317, 14327, 14478, 14496,
2978 14582, 14686, 14812, 14888, 14894, 14907, 14908, 14909, 14920, 14941,
2979 14952, 14964, 14981, 14982, 14985, 14991, 14994, 14996, 15000, 15003,
2980 15006, 15007, 15014, 15020, 15022, 15023, 15036, 15054, 15055, 15062,
2981 15078, 15084, 15085, 15086, 15100, 15160, 15214, 15221, 15232, 15234,
2982 15283, 15285, 15287, 15304, 15305, 15307, 15309, 15327, 15330, 15335,
2983 15336, 15337, 15339, 15342, 15346, 15359, 15361, 15366, 15380, 15381,
2984 15394, 15395, 15405, 15406, 15409, 15416, 15418, 15419, 15423, 15424,
2985 15426, 15429, 15431, 15432, 15441, 15442, 15448, 15465, 15480, 15485,
2986 15488, 15490, 15492, 15493, 15497, 15506, 15529, 15536, 15553, 15577,
2987 15583, 15618, 15627, 15631, 15654, 15655, 15666, 15667, 15674, 15711,
2988 15755, 15759.
2989
2990 * CVE-2013-2207 Incorrectly granting access to another user's pseudo-terminal
2991 has been fixed by disabling the use of pt_chown (Bugzilla #15755).
2992 Distributions can re-enable building and using pt_chown via the new configure
2993 option `--enable-pt_chown'. Enabling the use of pt_chown carries with it
2994 considerable security risks and should only be used if the distribution
2995 understands and accepts the risks.
2996
2997 * CVE-2013-0242 Buffer overrun in regexp matcher has been fixed (Bugzilla
2998 #15078).
2999
3000 * CVE-2013-1914 Stack overflow in getaddrinfo with many results has been
3001 fixed (Bugzilla #15330).
3002
3003 * Add support for calling C++11 thread_local object destructors on thread
3004 and program exit. This needs compiler support for offloading C++11
3005 destructor calls to glibc.
3006
3007 * Improved worst case performance of libm functions with double inputs and
3008 output.
3009
3010 * Support for priority inherited mutexes in pthread condition variables on
3011 non-x86 architectures.
3012
3013 * Port to Xilinx MicroBlaze contributed by David Holsgrove.
3014
3015 * Optimized string functions for AArch64. Implemented by Marcus Shawcroft.
3016
3017 * Optimized string functions for ARM. Implemented by Will Newton and
3018 Richard Henderson.
3019
3020 * Optimized libm functions for SPARC. Implemented by David S. Miller.
3021
3022 * Support for building more of ARM glibc as Thumb-2 code. Implemented by
3023 Richard Henderson.
3024
3025 * Support for building most of MIPS glibc as MIPS16 code. Implemented by
3026 Chung-Lin Tang, Maciej W. Rozycki and Maxim Kuvyrkov.
3027
3028 * Added a benchmark framework to track performance of functions in glibc.
3029
3030 * New <math.h> macro named issignaling to check for a signaling NaN (sNaN).
3031 It is based on draft TS 18661 and currently enabled as a GNU extension.
3032
3033 * On Linux, the clock function now uses the clock_gettime system call
3034 for improved precision, rather than old times system call.
3035
3036 * Added support for version-3 tz format files. This is needed when using
3037 the tz database release 2013e or later, and affects a few unusual cases --
3038 currently only TZ='America/Godthab' for time stamps after 2037.
3039
3040 * Added new API functions pthread_getattr_default_np and
3041 pthread_setattr_default_np to get and set the default pthread
3042 attributes of a process.
3043
3044 * Added support for TSX lock elision for pthread mutexes on i386 and x86-64.
3045 This may improve lock scaling of existing programs on TSX capable systems.
3046 When the --enable-lock-elision=yes parameter is specified at configure
3047 time lock elision will be enabled by default for all PTHREAD_MUTEX_DEFAULT
3048 mutexes.
3049
3050 * Support added for AT_HWCAP2 (to coincide with Linux kernel 3.10
3051 availability). Implemented by Ryan S. Arnold.
3052
3053 * Support added for POWER8 platform. Implemented by Ryan S. Arnold.
3054 \f
3055 Version 2.17
3056
3057 * The following bugs are resolved with this release:
3058
3059 1349, 3439, 3479, 3665, 5044, 5246, 5298, 5400, 6530, 6677, 6778, 6808,
3060 9685, 9914, 10014, 10038, 10114, 10191, 10631, 10873, 11438, 11607, 11638,
3061 11741, 12140, 13013, 13412, 13542, 13601, 13603, 13604, 13629, 13679,
3062 13696, 13698, 13717, 13741, 13759, 13761, 13763, 13881, 13939, 13950,
3063 13952, 13966, 14042, 14047, 14090, 14150, 14151, 14152, 14154, 14157,
3064 14166, 14173, 14195, 14197, 14237, 14246, 14251, 14252, 14283, 14298,
3065 14303, 14307, 14328, 14331, 14336, 14337, 14347, 14349, 14368, 14376,
3066 14417, 14447, 14459, 14476, 14477, 14501, 14505, 14510, 14516, 14518,
3067 14519, 14530, 14532, 14538, 14543, 14544, 14545, 14557, 14562, 14568,
3068 14576, 14579, 14583, 14587, 14595, 14602, 14610, 14621, 14638, 14645,
3069 14648, 14652, 14660, 14661, 14669, 14672, 14683, 14694, 14716, 14719,
3070 14743, 14767, 14783, 14784, 14785, 14793, 14796, 14797, 14801, 14803,
3071 14805, 14807, 14811, 14815, 14821, 14822, 14824, 14828, 14831, 14833,
3072 14835, 14838, 14856, 14863, 14865, 14866, 14868, 14869, 14871, 14872,
3073 14879, 14889, 14893, 14898, 14914.
3074
3075 * Optimization of memcpy for MIPS.
3076
3077 * CVE-2011-4609 svc_run() produces high cpu usage when accept fails with
3078 EMFILE has been fixed (Bugzilla #14889).
3079
3080 * The add-on ports collection is now distributed in the "ports" subdirectory
3081 of the main GNU C Library distribution, rather than separately.
3082
3083 * Port to ARM AArch64 contributed by Linaro.
3084
3085 * Support for STT_GNU_IFUNC symbols added for s390 and s390x.
3086 Optimized versions of memcpy, memset, and memcmp added for System z10 and
3087 zEnterprise z196.
3088 Implemented by Andreas Krebbel.
3089
3090 * The new function secure_getenv allows secure access to the environment,
3091 returning NULL if running in a SUID/SGID process. This function replaces
3092 the internal function __secure_getenv.
3093
3094 * SystemTap static probes have been added into the dynamic linker.
3095 Implemented by Gary Benson.
3096
3097 * Optimizations of string functions strstr, strcasestr and memmem.
3098 Implemented by Maxim Kuvyrkov.
3099
3100 * The minimum Linux kernel version that this version of the GNU C Library
3101 can be used with is 2.6.16.
3102
3103 * Optimizations of string functions memchr, wcschr, wcscpy, and wcsrchr for
3104 powerpc POWER7. Implemented by Will Schmidt.
3105
3106 * New configure option --disable-nscd builds the C library such that it
3107 never attempts to contact the Name Service Caching Daemon (nscd).
3108 New configure option --disable-build-nscd avoids building nscd itself;
3109 this is the default if --disable-nscd is used.
3110
3111 * Improved support for cross-compilation, including cross-testing and
3112 bootstrap builds without a previously built glibc.
3113
3114 * Several testsuite tests are now able to test multiple IFUNC variants of an
3115 interface, rather than just testing the one that would be chooen by
3116 default.
3117
3118 * New configure options --with-bugurl and --with-pkgversion, for
3119 distributors to use to embed their bug-reporting and package version
3120 information in --help and --version output.
3121
3122 * The ttyname and ttyname_r functions on Linux now fall back to searching for
3123 the tty file descriptor in /dev/pts or /dev if /proc is not available. This
3124 allows creation of chroots without the procfs mounted on /proc.
3125
3126 * The `crypt' function now fails if passed salt bytes that violate the
3127 specification for those values. On Linux, the `crypt' function will
3128 consult /proc/sys/crypto/fips_enabled to determine if "FIPS mode" is
3129 enabled, and fail on encrypted strings using the MD5 or DES algorithm
3130 when the mode is enabled.
3131
3132 * The `clock_*' suite of functions (declared in <time.h>) is now available
3133 directly in the main C library. Previously it was necessary to link with
3134 -lrt to use these functions. This change has the effect that a
3135 single-threaded program that uses a function such as `clock_gettime' (and
3136 is not linked with -lrt) will no longer implicitly load the pthreads
3137 library at runtime and so will not suffer the overheads associated with
3138 multi-thread support in other code such as the C++ runtime library.
3139
3140 * New locales: ayc_PE, doi_IN, ia_FR, mni_IN, nhn_MX, niu_NU, niu_NZ,
3141 sat_IN, and szl_PL.
3142 \f
3143 Version 2.16
3144
3145 * The following bugs are resolved with this release:
3146
3147 174, 208, 350, 369, 411, 706, 766, 2074, 2541, 2547, 2548, 2550, 2551,
3148 2552, 2553, 2554, 2562, 2563, 2565, 2566, 2570, 2576, 2636, 2678, 3335,
3149 3440, 3748, 3768, 3866, 3868, 3906, 3976, 3992, 4026, 4108, 4596, 4822,
3150 5077, 5461, 5805, 5993, 6471, 6486, 6578, 6649, 6730, 6770, 6794, 6884,
3151 6890, 6894, 6895, 6907, 6911, 6959, 7064, 9739, 9902, 10110, 10135, 10140,
3152 10153, 10210, 10254, 10346, 10375, 10545, 10716, 10846, 10882, 11174,
3153 11261, 11322, 11365, 11451, 11494, 11521, 11677, 11837, 11959, 12047,
3154 12097, 12193, 12194, 12297, 12298, 12301, 12340, 12354, 12416, 12495,
3155 13058, 13223, 13361, 13525, 13526, 13527, 13528, 13529, 13530, 13531,
3156 13532, 13533, 13547, 13551, 13552, 13553, 13555, 13556, 13559, 13563,
3157 13566, 13576, 13579, 13583, 13592, 13594, 13613, 13618, 13637, 13656,
3158 13658, 13673, 13691, 13695, 13704, 13705, 13706, 13718, 13726, 13738,
3159 13739, 13743, 13750, 13758, 13760, 13761, 13775, 13786, 13787, 13792,
3160 13806, 13824, 13840, 13841, 13844, 13846, 13848, 13851, 13852, 13854,
3161 13871, 13872, 13873, 13879, 13882, 13883, 13884, 13885, 13886, 13892,
3162 13895, 13908, 13910, 13911, 13912, 13913, 13914, 13915, 13916, 13917,
3163 13918, 13919, 13920, 13921, 13922, 13923, 13924, 13926, 13927, 13928,
3164 13938, 13941, 13942, 13954, 13955, 13956, 13963, 13967, 13968, 13970,
3165 13973, 13979, 13983, 13986, 13996, 14012, 14027, 14033, 14034, 14036,
3166 14040, 14043, 14044, 14048, 14049, 14050, 14053, 14055, 14059, 14064,
3167 14075, 14080, 14083, 14103, 14104, 14109, 14112, 14117, 14122, 14123,
3168 14134, 14153, 14183, 14188, 14199, 14210, 14218, 14229, 14241, 14273,
3169 14277, 14278.
3170
3171 * Support for the x32 ABI on x86-64 added. The x32 target is selected by
3172 configuring glibc with:
3173 BUILD_CC='gcc' CC='gcc -mx32' CXX='g++ -mx32'
3174 Visit <http://sites.google.com/site/x32abi/> for more x32 ABI info.
3175 Implemented by H.J. Lu.
3176
3177 * ISO C11 support:
3178
3179 + define static_assert
3180
3181 + do not declare gets
3182
3183 + declare at_quick_exit and quick_exit also for ISO C11
3184
3185 + aligned_alloc. NB: The code is deliberately allows the size parameter
3186 to not be a multiple of the alignment. This is a moronic requirement
3187 in the standard but it is only a requirement on the caller, not the
3188 implementation.
3189
3190 + timespec_get added
3191
3192 + uchar.h support added
3193
3194 + CMPLX, CMPLXF, CMPLXL added
3195
3196 Implemented by Ulrich Drepper.
3197
3198 * Support for the IA-64 has been moved to ports.
3199
3200 * Remove support for anything but ELF binary format
3201
3202 * Checking versions of poll, ppoll added.
3203 Implemented by Ulrich Drepper.
3204
3205 * More generic and 64-bit performance optimizations to math functions.
3206 Implemented by Ulrich Drepper.
3207
3208 * New configure option --enable-obsolete-rpc makes the deprecated RPC
3209 headers and functions available at compile time as they were before
3210 version 2.14. This option will be removed at some time in the future
3211 after the TI-RPC library becomes fully sufficient for the needs of
3212 existing applications.
3213
3214 * Compatibility code for Linux kernel versions before 2.4 has been removed.
3215 Note that glibc is not expected to work with any Linux kernel version
3216 before 2.6.
3217
3218 * New header <sys/auxv.h> and function getauxval allowing easy access to
3219 the AT_* key-value pairs passed from the Linux kernel. The header also
3220 defines the HWCAP_* bits associated with the AT_HWCAP key.
3221
3222 * New locales: mag_IN
3223
3224 * New configure option --enable-systemtap builds SystemTap static probes
3225 into libc for setjmp and longjmp and into libpthread for various operations.
3226 So far the setjmp/longjmp probes and some of the libpthread probes are
3227 provided only for i*86 and x86_64.
3228 Implemented by Roland McGrath and Rayson Ho.
3229
3230 * Optimized expf for x86-32 and x86-64. Implemented by Liubov Dmitrieva.
3231
3232 * More optimized functions for PowerPC. Implemented by Adhemerval Zanella
3233 and Will Schmidt.
3234
3235 * More optimized functions for SPARC. Implemented by David S. Miller.
3236
3237 * Improved support for cross-compilation, especially bootstrap builds
3238 without a previously built glibc.
3239
3240 * Ports for the TILE-Gx and TILEPro families of processors. Contributed by
3241 Chris Metcalf from Tilera.
3242
3243 * Support for the old ARM ABI has been removed from ports. Only the EABI is
3244 now supported for ARM processors.
3245
3246 * The hard-float variant of the ARM EABI now uses /lib/ld-linux-armhf.so.3
3247 as the name of the dynamic linker, to distinguish it from the
3248 /lib/ld-linux.so.3 dynamic linker used for the base EABI.
3249
3250 * Support for CRIS has been removed from ports.
3251
3252 * A new class of installed header has been documented for low-level
3253 platform-specific functionality. PowerPC added the first instance with a
3254 function to provide time base register access. Contributed by Tulio
3255 Magno Quites Machado Filho.
3256
3257 * ix86 configurations now install header files that are consistent with
3258 what x86-64 configurations install. These same header files can be used
3259 for -m32, -m64, or -mx32 builds.
3260 Contributed by H.J. Lu.
3261
3262 * Math library bug fixes. A thorough audit of all open math library bugs was
3263 conducted by Joseph Myers. Significant progress was made on many math
3264 library bugs resulting in more accurate exceptions and function results.
3265 Many thanks to all those that contributed including Andreas Jaeger for his
3266 patch review and work on the x87 trigonometric instruction issues.
3267
3268 * Timezone data is no longer installed. Timezone-related binaries and scripts
3269 will continue to be installed. Users should obtain their timezone data from
3270 their distribution provider or from the tzdata package at
3271 <ftp://munnari.oz.au/pub/>.
3272 \f
3273 Version 2.15
3274
3275 * The following bugs are resolved with this release:
3276
3277 6779, 6783, 9696, 10103, 10709, 11589, 11929, 12403, 12786, 12840, 12847,
3278 12868, 12852, 12874, 12885, 12892, 12906, 12907, 12922, 12935, 12962,
3279 13007, 13021, 13061, 13062, 13067, 13068, 13085, 13088, 13090, 13092,
3280 13096, 13114, 13118, 13123, 13134, 13138, 13147, 13150, 13166, 13179,
3281 13185, 13189, 13192, 13268, 13276, 13282, 13291, 13305, 13328, 13335,
3282 13337, 13344, 13358, 13367, 13413, 13416, 13423, 13439, 13446, 13472,
3283 13484, 13506, 13515, 13523, 13524, 13538, 13540
3284
3285 * New program pldd to list loaded object of a process
3286 Implemented by Ulrich Drepper.
3287
3288 * Add nss_db support back to glibc. No more dependency on Berkeley db
3289 and support for initgroups lookups.
3290 Implemented by Ulrich Drepper.
3291
3292 * Optimized strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-32.
3293 Contributed by HJ Lu.
3294
3295 * Improved strcpy, strncpy, stpcpy, stpncpy for SSE2 and SSSE3 on x86-64.
3296 Contributed by HJ Lu.
3297
3298 * Optimized strcat, strncat on x86-64 and optimized wcscmp, wcslen, strnlen
3299 on x86-32 and x86-64.
3300 Contributed by Liubov Dmitrieva.
3301
3302 * Optimized strchr and strrchr for SSE on x86-32.
3303 Contributed by Liubov Dmitrieva.
3304
3305 * Optimized memchr, memrchr, rawmemchr, memcmp, wmemcmp, wcschr, wcscpy
3306 for x86-64 and x86-32.
3307 Contributed by Liubov Dmitrieva.
3308
3309 * New interfaces: scandirat, scandirat64
3310 Implemented by Ulrich Drepper.
3311
3312 * Checking versions of FD_SET, FD_CLR, and FD_ISSET added.
3313 Implemented by Ulrich Drepper.
3314
3315 * nscd now also caches the netgroup database.
3316 Implemented by Ulrich Drepper.
3317
3318 * Integrate libm with gcc's -ffinite-math-only option.
3319 Implemented by Ulrich Drepper.
3320
3321 * Lots of generic, 64-bit, and x86-64-specific performance optimizations
3322 to math functions. Implemented by Ulrich Drepper.
3323
3324 * Optimized strcasecmp and strncasecmp for AVX on x86-64.
3325 Implemented by Ulrich Drepper.
3326
3327 * New Linux interfaces: process_vm_readv, process_vm_writev
3328
3329 * Optimized strcasecmp and strncasecmp for SSSE3 and SSE4.2 on x86-32.
3330 Implemented by Ulrich Drepper.
3331
3332 * Optimized nearbyint and strcasecmp for PPC.
3333 Implemented by Adhemerval Zanella.
3334
3335 * New locales: bho_IN, unm_US, es_CU, ta_LK
3336 \f
3337 Version 2.14
3338
3339 * The following bugs are resolved with this release:
3340
3341 386, 6420, 7101, 9730, 9732, 9809, 10138, 10149, 10157, 11099, 11257,
3342 11258, 11487, 11532, 11558, 11578, 11634, 11653, 11668, 11697, 11724,
3343 11781, 11799, 11820, 11837, 11857, 11884, 11892, 11895, 11901, 11945,
3344 11947, 11952, 11987, 12052, 12083, 12158, 12178, 12200, 12346, 12350,
3345 12393, 12420, 12432, 12445, 12449, 12453, 12454, 12460, 12469, 12489,
3346 12509, 12510, 12511, 12518, 12527, 12541, 12545, 12551, 12582, 12583,
3347 12587, 12597, 12601, 12611, 12625, 12626, 12631, 12650, 12653, 12655,
3348 12660, 12671, 12681, 12684, 12685, 12711, 12713, 12714, 12717, 12723,
3349 12724, 12734, 12738, 12746, 12766, 12775, 12777, 12782, 12788, 12792,
3350 12795, 12811, 12813, 12814, 12841
3351
3352 * The RPC implementation in libc is obsoleted. Old programs keep working
3353 but new programs cannot be linked with the routines in libc anymore.
3354 Programs in need of RPC functionality must be linked against TI-RPC.
3355 The TI-RPC implementation is IPv6 enabled and there are other benefits.
3356
3357 Visible changes of this change include (obviously) the inability to link
3358 programs using RPC functions without referencing the TI-RPC library and the
3359 removal of the RPC headers from the glibc headers.
3360 Implemented by Ulrich Drepper.
3361
3362 * New Linux interfaces: clock_adjtime, name_to_handle_at, open_by_handle_at,
3363 syncfs, setns, sendmmsg
3364
3365 * New locales: os_RU, bem_ZA, en_ZA, ff_SN, sw_KE, sw_TZ, lb_LU, wae_CH,
3366 yue_HK, lij_IT, mhr_RU
3367
3368 * New encodings: CP770, CP771, CP772, CP773, CP774
3369
3370 * New program sotruss to trace calls through PLTs
3371 Implemented by Ulrich Drepper.
3372
3373 * The malloc hook implementation is marked deprecated and will be removed
3374 from the default implementation in the next version. The design never
3375 worked ever since the introduction of threads. Even programs which do
3376 not create threads themselves can use multiple threads created internally.
3377 \f
3378 Version 2.13
3379
3380 * The following bugs are resolved with this release:
3381
3382 3268, 6812, 7066, 10085, 10484, 10563, 10851, 11149, 11155, 11577, 11611,
3383 11640, 11655, 11701, 11840, 11856, 11883, 11903, 11904, 11968, 11979,
3384 12005, 12037, 12067, 12077, 12078, 12092, 12093, 12107, 12108, 12113,
3385 12140, 12159, 12167, 12191, 12194, 12201, 12204, 12205, 12207, 12348,
3386 12378, 12394, 12397
3387
3388 * New Linux interfaces: prlimit, prlimit64, fanotify_init, fanotify_mark
3389
3390 * POWER7 optimizations: memset, memcmp, strncmp
3391
3392 * New optimized string functions for x86-64: strnlen (SSE2),
3393 strcasecmp (SSE2, SSSE3, SSE4.2), strncasecmp (SSE2, SSSE3, SSE4.2)
3394 Implemented by Ulrich Drepper.
3395 \f
3396 Version 2.12
3397
3398 * The following bugs are resolved with this release:
3399
3400 3662, 4457, 5553, 10162, 10401, 10414, 10554, 10824, 10864, 10915,
3401 10918, 10936, 10939, 10958, 10968, 10969, 10972, 10992, 11000, 11001,
3402 11007, 11010, 11027, 11039, 11040, 11041, 11043, 11046, 11056, 11070,
3403 11093, 11115, 11125, 11126, 11127, 11134, 11141, 11149, 11183, 11184,
3404 11185, 11186, 11187, 11188, 11189, 11190, 11191, 11192, 11193, 11194,
3405 11200, 11230, 11235, 11242, 11254, 11258, 11271, 11272, 11276, 11279,
3406 11287, 11292, 11319, 11332, 11333, 11387, 11389, 11390, 11394, 11397,
3407 11410, 11438, 11449, 11470, 11471, 11520, 11537, 11538, 11571
3408
3409 * New interfaces: pthread_getname_np, pthread_setname_np
3410
3411 * New Linux interface: recvmmsg
3412
3413 * STT_GNU_IFUNC implemented for Sparc by David Miller.
3414
3415 * The dynamic linker now recognizes supported ABI versions from the
3416 EI_ABIVERSION field in the ELF header.
3417 Implemented by Ulrich Drepper.
3418
3419 * New NIS mode selector ADJUNCT_AS_SHADOW. The passwd.adjunct.byname table
3420 will not be used to fill in password fields in the passwd.byname replies.
3421 Instead it is used to synthesize the shadow.byname table, should it be
3422 missing. This is a useful mode in some installations involving Solaris.
3423 Implemented by Ulrich Drepper.
3424
3425 * New locales: kok_IN, sq_MK, cv_RU
3426 \f
3427 Version 2.11
3428
3429 * The following bugs are resolved with this release:
3430
3431 3270, 7094, 9924, 9986, 10011, 10085, 10107, 10159, 10162, 10166, 10183,
3432 10195, 10196, 10200, 10203, 10207, 10211, 10213, 10217, 10221, 10229,
3433 10262, 10286, 10312, 10315, 10319, 10349, 10360, 10391, 10402, 10416,
3434 10418, 10422, 10425, 10446, 10448, 10452, 10540, 10553, 10560, 10564,
3435 10609, 10643, 10692, 10713, 10717, 10730, 10731, 10742, 10780, 10784,
3436 10789, 10817, 10823, 10840, 10847
3437
3438 * New interfaces: execvpe, pthread_sigqueue, mkstemps, mkstemps64, mkostemps,
3439 mkostemps64
3440 Implemented by Ulrich Drepper.
3441
3442 * Checking version of longjmp added that fails if an uninitialized stack
3443 frame would be created. Implemented by Ulrich Drepper.
3444
3445 * STT_GNU_IFUNC is now supported in static executables.
3446 Implemented by H.J. Lu.
3447
3448 * STT_GNU_IFUNC implemented for PPC by Alan Modra.
3449
3450 * New optimized string functions for x86-64: strstr, strcasestr, memcmp,
3451 strcspn, strpbrk, strspn, strcpy, stpcpy, strncpy, strcmp (SSE2, SSE4.2),
3452 strncmp (SSE2, SSE4.2), strchr (SSE4.2), strrchr (SSE4.2).
3453 Contributed by H.J. Lu.
3454
3455 strlen, rawmemchr, strcmp (SSSE3), strncmp (SSSE3).
3456 Implemented by Ulrich Drepper.
3457
3458 * New optimized string functions for x86: strlen, strcspn, strspn, strpbrk,
3459 strstr, strcasestr.
3460 Contributed by H.J. Lu.
3461
3462 * Support for fma instruction in AVX on x86-64.
3463 Implemented by H.J. Lu and Ulrich Drepper.
3464
3465 * AVX support in x86-64 auditing support in ld.so.
3466 Implemented by H.J. Lu.
3467
3468 * STB_GNU_UNIQUE support added.
3469 Implemented by Ulrich Drepper.
3470
3471 * Implement second fallback mode for optimized DNS lookup for even more
3472 broken environments. If two requests from the same port are not
3473 handled correctly close the socket and open a new one before sending
3474 the second request. The 'single-request-reopen' option in /etc/resolv.conf
3475 can be used to select this mode right away, instead of rediscovering the
3476 necessity is every process again.
3477 Implemented by Ulrich Drepper.
3478
3479 * New resolver flag RES_USE_DNSSEC to enable use of verified lookup.
3480 Implemented by Adam Tkac.
3481
3482 * Optimized iconv conversions for S390x.
3483 Implemented by Andreas Krebbel.
3484
3485 * Using condvars with PI mutexes is now more efficient due to kernel
3486 support for requeueing to PI futexes. NPTL support added for x86-64.
3487 Implemented by Ulrich Drepper.
3488
3489 * New locales: ps_AF, my_MM
3490 \f
3491 Version 2.10
3492
3493 * The following bugs are resolved with this release:
3494
3495 697, 5381, 5807, 6411, 6545, 7040, 7058, 7067, 7080, 7083, 7095, 7098,
3496 9677, 9694, 9697, 9701, 9704, 9705, 9706, 9720, 9726, 9730, 9733, 9736,
3497 9741, 9750, 9753, 9759, 9781, 9791, 9793, 9823, 9833, 9844, 9880, 9881,
3498 9891, 9893, 9895, 9913, 9916, 9920, 9948, 9955, 9957, 9963, 9985, 10052,
3499 10069, 10086, 10087, 10090, 10092, 10093, 10100, 10118, 10128
3500
3501 * New interfaces: preadv, preadv64, pwritev, pwritev64, malloc_info
3502 Implemented by Ulrich Drepper.
3503
3504 * New Linux interfaces: accept4, fallocate, fallocate64.
3505 Implemented by Ulrich Drepper.
3506
3507 * Correct declarations of string function when used in C++ code. This
3508 could lead to compile errors for invalid C++ code.
3509
3510 * XPG7/POSIX 2008 compilation environment. Many GNU-specific functions are
3511 now in POSIX.
3512
3513 * New POSIX 2008 interface: psiginfo
3514 Implemented by Ulrich Drepper.
3515
3516 * New ISO C++1x interfaces: quick_exit, at_quick_exit
3517 Implemented by Ulrich Drepper.
3518
3519 * Support for selecting between multiple function definitions at runtime
3520 using STT_GNU_IFUNC symbols. Implemented by Ulrich Drepper.
3521
3522 * The libcrypt library can now use the hash function implementations in
3523 NSS. Implemented by Ulrich Drepper.
3524
3525 * The malloc implementation can be compiled to be less memory efficient
3526 but higher performing in multi-threaded programs.
3527 Implemented by Ulrich Drepper.
3528
3529 * New locales: nan_TW@latin, ks_IN
3530
3531 * Faster strlen, strchr, strchrnul, strrchr, memchr, and rawmemchr for x86-64.
3532 Implemented by Ulrich Drepper.
3533
3534 * Extended printf hook support. It is possible to use user-defined types
3535 and extend existing format specifiers.
3536 Implemented by Ulrich Drepper.
3537
3538 * Handling for group shadow files has been added.
3539 Implemented by Ulrich Drepper.
3540
3541 * DNS IPv4-IPv6 parallel lookup now deals better with broken DNS
3542 servers (the case, e.g., for some people using the built-in DNS
3543 server in ADSL modems/routers). There is a once-per-process timeout
3544 in case of a broken server. To avoid it, users can run nscd or put
3545 'options single-request' in /etc/resolv.conf.
3546 Implemented by Ulrich Drepper.
3547 \f
3548 Version 2.9
3549
3550 * The following bugs are resolved with this release:
3551
3552 3406, 5209, 5210, 5381, 5794, 5814, 5911, 6428, 6438, 6442, 6447, 6461,
3553 6472, 6505, 6544, 6589, 6612, 6634, 6653, 6654, 6657, 6698, 6712, 6713,
3554 6719, 6723, 6724, 6740, 6763, 6771, 6790, 6791, 6817, 6824, 6839, 6843,
3555 6845, 6856, 6857, 6860, 6867, 6875, 6919, 6920, 6942, 6947, 6955, 6966,
3556 6968, 6974, 6980, 6995, 7008, 7009, 7029
3557
3558 * Unified lookup for getaddrinfo: IPv4 and IPv6 addresses are now looked
3559 up at the same time. Implemented by Ulrich Drepper.
3560
3561 * TLS descriptors for LD and GD on x86 and x86-64.
3562 Implemented by Alexandre Oliva.
3563
3564 * getaddrinfo now handles DCCP and UDPlite.
3565 Implemented by Ulrich Drepper.
3566
3567 * New fixed-size conversion macros: htobe16, htole16, be16toh, le16toh,
3568 htobe32, htole32, be32toh, le32toh, htobe64, htole64, be64toh, le64toh.
3569 Implemented by Ulrich Drepper.
3570
3571 * New implementation of memmem, strstr, and strcasestr which is O(n).
3572 Implemented by Eric Blake.
3573
3574 * New Linux interfaces: inotify_init1, dup3, epoll_create1, pipe2
3575
3576 * Implement "e" option for popen to open file descriptor with the
3577 close-on-exec flag set. Implemented by Ulrich Drepper.
3578
3579 * Implement "b" mode for fmemopen. In this mode writes writes don't
3580 implicitly add a NUL byte and seeks from the end of the buffer really
3581 use the buffer end, not the string length as the basis.
3582 Implemented by Ulrich Drepper.
3583
3584 * Many functions, exported and internal, now atomically set the close-on-exec
3585 flag when run on a sufficiently new kernel. Implemented by Ulrich Drepper.
3586
3587 * Sorting rules for some Indian languages (Kannada, Gurumukhi, Telugu,
3588 Sinhala)
3589 Implemented by Pravin Satpute.
3590
3591 * New locales: sd_IN, sd_IN@devanagari, ks_IN@devanagari, ht_HT, en_AG, nl_AW.
3592 \f
3593 Version 2.8
3594
3595 * The following bugs are resolved with this release:
3596
3597 2549, 3112, 4314, 4407, 4997, 5012, 5040, 5112, 5204, 5208, 5209, 5220,
3598 5222, 5224, 5225, 5237, 5238, 5240, 5245, 5277, 5346, 5375, 5378, 5382,
3599 5424, 5427, 5428, 5435, 5436, 5439, 5441, 5442, 5443, 5451, 5452, 5454,
3600 5455, 5463, 5464, 5465, 5475, 5477, 5478, 5520, 5541, 5545, 5553, 5600,
3601 5602, 5607, 5614, 5627, 5628, 5736, 5737, 5741, 5753, 5760, 5762, 5768,
3602 5774, 5777, 5778, 5779, 5786, 5790, 5818, 5854, 5857, 5903, 5939, 5979,
3603 5983, 5995, 5998, 6004, 6007, 6020, 6021, 6024, 6040, 6042
3604
3605 * New locales: bo_CN, bo_IN, shs_CA.
3606
3607 * New encoding: HP-ROMAN9, HP-GREEK8, HP-THAI8, HP-TURKISH8.
3608
3609 * Sorting rules for some Indian languages (Devanagari and Gujarati).
3610 Implemented by Pravin Satpute.
3611
3612 * IPV6 addresses in /etc/resolv.conf can now have a scope ID
3613
3614 * nscd caches now all timeouts for DNS entries
3615 Implemented by Ulrich Drepper.
3616
3617 * nscd is more efficient and wakes up less often.
3618 Implemented by Ulrich Drepper.
3619
3620 * More checking functions: asprintf, dprintf, obstack_printf, vasprintf,
3621 vdprintf, and obstack_vprintf.
3622 Implemented by Jakub Jelinek.
3623
3624 * Faster memset for x86-64.
3625 Implemented by Harsha Jagasia and H.J. Lu.
3626
3627 * Faster memcpy on x86.
3628 Implemented by Ulrich Drepper.
3629
3630 * ARG_MAX is not anymore constant on Linux. Use sysconf(_SC_ARG_MAX).
3631 Implemented by Ulrich Drepper.
3632
3633 * Faster sqrt and sqrtf implemention for some PPC variants.
3634 Implemented by Steven Munroe.
3635 \f
3636 Version 2.7
3637
3638 * The following bugs are resolved with this release:
3639
3640 4125, 4126, 4178, 4359, 4407, 4512, 4514, 4525, 4554, 4556, 4557, 4566,
3641 4582, 4586, 4588, 4599, 4610, 4647, 4702, 4705, 4726, 4745, 4772, 4773,
3642 4775, 4776, 4792, 4813, 4814, 4816, 4833, 4858, 4860, 4896, 4905, 4925,
3643 4936, 4937, 4938, 4941, 4946, 4963, 4972, 5010, 5028, 5043, 5058, 5063,
3644 5071, 5103, 5104, 5112, 5113, 5184, 5186
3645
3646 * More checking functions: fread, fread_unlocked, open*, mq_open.
3647 Implemented by Jakub Jelinek and Ulrich Drepper.
3648
3649 * Extend fortification to C++. Implemented by Jakub Jelinek.
3650
3651 * Implement 'm' modifier for scanf. Add stricter C99/SUS compliance
3652 by not recognizing 'a' as a modifier when those specs are requested.
3653 Implemented by Jakub Jelinek.
3654
3655 * PPC optimizations to math and string functions.
3656 Implemented by Steven Munroe.
3657
3658 * New interfaces: mkostemp, mkostemp64. Like mkstemp* but allow additional
3659 options to be passed. Implemented by Ulrich Drepper.
3660
3661 * More CPU set manipulation functions. Implemented by Ulrich Drepper.
3662
3663 * New Linux interfaces: signalfd, eventfd, eventfd_read, and eventfd_write.
3664 Implemented by Ulrich Drepper.
3665
3666 * Handle private futexes in the NPTL implementation.
3667 Implemented by Jakub Jelinek and Ulrich Drepper.
3668
3669 * Add support for O_CLOEXEC. Implement in Hurd. Use throughout libc.
3670 Implemented by Roland McGrath and Ulrich Drepper.
3671
3672 * Linux/x86-64 vDSO support. Implemented by Ulrich Drepper.
3673
3674 * SHA-256 and SHA-512 based password encryption.
3675 Implemented by Ulrich Drepper.
3676
3677 * New locales: ber_DZ, ber_MA, en_NG, fil_PH, fur_IT, fy_DE, ha_NG, ig_NG,
3678 ik_CA, iu_CA, li_BE, li_NL, nds_DE, nds_NL, pap_AN, sc_IT, tk_TM, ug_CN,
3679 yo_NG.
3680
3681 + New iconv modules: MAC-CENTRALEUROPE, ISO-8859-9E, KOI8-RU.
3682 Implemented by Ulrich Drepper.
3683 \f
3684 Version 2.6
3685
3686 * The following bugs are resolved with this release:
3687
3688 3156, 3213, 3285, 3291, 3306, 3313, 3320, 3322, 3325, 3326, 3334, 3348,
3689 3352, 3363, 3369, 3426, 3427, 3429, 3451, 3458, 3465, 3480, 3483, 3493,
3690 3514, 3515, 3559, 3632, 3664, 3673, 3674, 3745, 3747, 3818, 3840, 3842,
3691 3851, 3855, 3884, 3885, 3902, 3905, 3919, 3944, 3954, 3955, 3957, 3991,
3692 3995, 3996, 4024, 4040, 4069, 4070, 4074, 4076, 4096, 4101, 4102, 4114,
3693 4130, 4131, 4181, 4276, 4306, 4342, 4344, 4349, 4364, 4368, 4381, 4392,
3694 4403, 4405, 4406, 4411, 4438, 4439, 4455, 4456, 4465, 4512, 4514, 4586,
3695 4702, 4858
3696
3697 * New Linux interfaces: epoll_pwait, sched_getcpu.
3698
3699 * New generic interfaces: strerror_l.
3700
3701 * nscd can now cache the services database. Implemented by Ulrich Drepper.
3702 \f
3703 Version 2.5
3704
3705 * The following bugs are resolved with this release:
3706
3707 39, 68, 192, 352, 388, 531, 935, 1006, 1201, 1203, 1386, 1782, 1783, 1784,
3708 1785, 1786, 1787, 1951, 1974, 1996, 2072, 2098, 2125, 2134, 2158, 2182,
3709 2349, 2376, 2386, 2389, 2415, 2418, 2420, 2423, 2450, 2451, 2466, 2472,
3710 2473, 2477, 2498, 2499, 2501, 2502, 2505, 2507, 2508, 2509, 2517, 2526,
3711 2569, 2571, 2592, 2611, 2625, 2632, 2644, 2662, 2680, 2683, 2684, 2693,
3712 2695, 2703, 2734, 2751, 2764, 2766, 2775, 2792, 2821, 2823, 2832, 2841,
3713 2843, 2883, 2892, 2908, 2914, 2926, 2961, 2978, 2980, 2987, 2997, 2998,
3714 3013, 3018, 3034, 3040, 3044, 3123, 3124, 3137, 3138, 3143, 3155, 3189,
3715 3225, 3251, 3252, 3253, 3273
3716
3717 * For Linux, the sorting of addresses returned by getaddrinfo now also
3718 handles rules 3, 4, and 7 from RFC 3484. I.e., all rules are handled.
3719 Implemented by Ulrich Drepper.
3720
3721 * Allow system admin to configure getaddrinfo with the /etc/gai.conf file.
3722 Implemented by Ulrich Drepper.
3723
3724 * New Linux interfaces: splice, tee, sync_file_range, vmsplice.
3725
3726 * New iconv module for MIK. Contributed by Alexander Shopov.
3727
3728 * For sites with broken group and/or passwd database, the auto-propagate
3729 option of nscd can prevent creating ID lookup entries from the results
3730 of a name lookup and vice versa. This usually is no problem but some
3731 site might have problems with the default behavior.
3732 Implemented by Ulrich Drepper.
3733
3734 * Iterating over entire database in NIS can be slow. With the
3735 SETENT_BATCH_READ option in /etc/default/nss a system admin can decide
3736 to trade time for memory. The entire database will be read at once.
3737 Implemented by Ulrich Drepper.
3738
3739 * The interfaces introduced in RFC 3542 have been implemented by
3740 Ulrich Drepper.
3741
3742 * Support for the new ELF hash table format was added by Ulrich Drepper.
3743
3744 * Support for priority inheritance mutexes added by Jakub Jelinek and
3745 Ulrich Drepper.
3746
3747 * Support for priority protected mutexes added by Jakub Jelinek.
3748
3749 * New locales: nr_ZA, pa_PK, ca_AD, ca_FR, ca_IT, el_CY, tr_CY, as_IN,
3750 or_IN, csb_PL, fy_NL, sr_ME.
3751 \f
3752 Version 2.4
3753
3754 * More overflow detection functions.
3755
3756 * New iconv converters for IBM1025, IBM1122, IBM1137, IBM1153,
3757 IBM1154, IBM1155, IBM1156, IBM1157, and IBM1158 contributed by Jiro Sekiba.
3758
3759 More converters for IBM803, IBM901, IBM902, IBM921, IBM1008, IBM1097,
3760 IBM1112, IBM1123, IBM1130, IBM1140, IBM1141, IBM1142, IBM1143, IBM1144,
3761 IBM1145, IBM1146, IBM1147, IBM1148, IBM1149, IBM1166, IBM1167, IBM4517,
3762 IBM4899, IBM4909, IBM4971, IBM5347, IBM9030, IBM9066, IBM9448, IBM12712,
3763 IBM16804, IBM1364, IBM1371, IBM1388, IBM1390, and IBM1399 contributed
3764 by Masahide Washizawa.
3765
3766 * It is now possible to install audit libraries for the dynamic linker using
3767 LD_AUDIT. Implemented by Ulrich Drepper.
3768
3769 * The LinuxThreads add-on, providing pthreads on Linux 2.4 kernels, is no
3770 longer supported. The new NPTL implementation requires Linux 2.6 kernels.
3771 For a libc and libpthread that works well on Linux 2.4 kernels, we
3772 recommend using the stable 2.3 branch.
3773
3774 * The new function open_wmemstream acts like open_memstream,
3775 but uses a wchar_t wide character buffer.
3776
3777 * The new function ppoll is an improved version of poll, similar to pselect.
3778
3779 * New interfaces: fdopendir, openat, fstatat, fchmodat, fchownat,
3780 futimesat, faccessat, mkdirat, mkfifoat, mknodat,
3781 renameat, unlinkat, linkat, symlinkat, readlinkat.
3782
3783 * New Linux kernel system calls: unshare,
3784 inotify_init, inotify_add_watch, inotify_rm_watch.
3785
3786 * The euidaccess function is now also known by the alias eaccess,
3787 for compatibility with some other systems.
3788
3789 * Timezone data updated to 2006b version.
3790 \f
3791 Version 2.3.6
3792
3793 * The following bugs are resolved with this release:
3794
3795 38, 253, 549, 622, 653, 721, 758, 851, 877, 915, 934, 955, 961,
3796 1016, 1037, 1076, 1079, 1080, 1081, 1082, 1083, 1084, 1085, 1086,
3797 1087, 1088, 1090, 1091, 1092, 1093, 1094, 1095, 1096, 1097, 1098,
3798 1099, 1100, 1101, 1102, 1103, 1104, 1105, 1106, 1107, 1108, 1109,
3799 1110, 1111, 1112, 1113, 1125, 1137, 1138, 1249, 1250, 1251, 1252,
3800 1253, 1254, 1350, 1358, 1394, 1438, 1498, 1534
3801
3802 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
3803
3804 * As of this release, GCC 4 can be used to compile the C Library.
3805
3806 * Timezone data updated to 2005m version.
3807 \f
3808 Version 2.3.5
3809
3810 * The following bugs are resolved with this release:
3811
3812 284, 592, 602, 626, 633, 640, 650, 661, 671, 681, 693, 700, 710, 719,
3813 722, 723, 725, 726, 727, 728, 729, 730, 731, 732, 733, 734, 735, 736,
3814 737, 738, 739, 740, 741, 742, 743, 744, 745, 765, 767, 768, 769, 776,
3815 777, 787, 821, 822, 823, 825
3816
3817 Visit <http://sources.redhat.com/bugzilla/> for the details of each bug.
3818 \f
3819 Version 2.3.4
3820
3821 * Support for RFC 3678. Real implementations exist only for Linux so far.
3822 Implemented by Ulrich Drepper.
3823
3824 * nscd can now cache entries persistently. Expiring entries are reloaded.
3825 For speedups the cache can be shared in memory with client processes.
3826 Implemented by Ulrich Drepper.
3827
3828 * nscd can now perform SELinux checks.
3829 Implemented by Matthew Rickard <mjricka@epoch.ncsc.mil>.
3830
3831 * getaddrinfo queries are now cached. Canonical name lookup is performed
3832 efficiently.
3833 Implemented by Ulrich Drepper.
3834
3835 * The nothrow function attribute is used when headers are used by gcc when
3836 compiling C code. This can avoid significant amounts of exception
3837 handling data.
3838
3839 * The malloc functions perform more error checking and are stricter when
3840 it comes to reacting on errors. The default action is to terminate
3841 the process after showing an error message. Implemented by Ulrich Drepper.
3842
3843 * Reverse lookups of IPv6 addresses does not use bit string or .ip6.int
3844 lookups anymore unless explicitly requested. Implemented by Ulrich Drepper.
3845
3846 * Namespaces in ld.so are implemented. DSOs can be loaded in separate
3847 namespaces using the new function dlmopen(). This feature is of course,
3848 like most other dynamic loading functionality, not available in statically
3849 linked applications. Implemented by Ulrich Drepper.
3850
3851 * Low-overhead boundary checking variants of string and some stdio functions
3852 were added. These are to be used in conjunction with a gcc patch by
3853 Jakub Jelinek which adds calls to these functions if possible.
3854 Implemented by Jakub Jelinek and Ulrich Drepper.
3855
3856 * Old code for several operating systems and machine architectures that
3857 have not been in working condition in a long time have been removed from
3858 the main source tree maintained by the GNU C Library's maintainers.
3859 These files are now reside in the separate `ports' source module
3860 that is usable as an add-on when building the library.
3861 \f
3862 Version 2.3.3
3863
3864 * New functions `dladdr1' and `dlinfo' in <dlfcn.h> provide more ways to
3865 interrogate the dynamic linker, compatible with the Solaris interface.
3866
3867 * ELF thread-local storage support (TLS) now works on PowerPC and PowerPC64;
3868 implemented by Paul Mackerras, Steven Munroe, and Roland McGrath.
3869
3870 * getifaddrs now uses the netlink interface on Linux to get its information.
3871 Implemented by Thorsten Kukuk.
3872
3873 * getaddrinfo now implements AI_V4MAPPED, AI_ALL, and AI_ADDRCONF.
3874 Implemented by Ulrich Drepper.
3875
3876 * support for non-executable stacks on x86 has been added. Changes mostly
3877 by Roland McGrath.
3878
3879 * regex is now much faster for multibyte locales. Changes by Jakub Jelinek
3880 and Ulrich Drepper.
3881
3882 * getaddrinfo now performs destination address selection according to
3883 RFC 3484.
3884 \f
3885 Version 2.3.2
3886
3887 * Thread-safe interfaces for many functions that access locale data
3888 were added in version 2.3, but these features were omitted from NEWS.
3889 Many functions have variants with an `_l' suffix that take a `locale_t'
3890 object as a parameter rather than consulting the current locale.
3891 The new functions `newlocale', `duplocale', and `freelocale' in <locale.h>
3892 create and maintain `locale_t' objects. Additionally, the new function
3893 `uselocale' sets "the current locale" (as used by functions not so
3894 parameterized) set for an individual thread. These features were added
3895 in version 2.3, implemented by Ulrich Drepper and Roland McGrath.
3896
3897 * The functions getresuid, getresgid, setresuid, and setresgid, which
3898 have long been available on Linux, are now declared in <unistd.h>
3899 and are now also available on the Hurd.
3900
3901 * ELF thread-local storage support (TLS) now works on x86-64.
3902
3903 * The new dynamic string token $LIB is expanded in shared library names.
3904 This normally expands to lib, but on some 64-bit platforms to lib64 instead.
3905
3906 * Aldy Hernandez contributed complete software floating point support for
3907 PowerPC machines with no FPU.
3908
3909 * fexecve is implemented on Linux.
3910
3911 * The `btowc' function should work at least twice as fast due to
3912 specialized callbacks in the iconv modules. Implemented by Bruno Haible.
3913
3914 * With approriate thread add-ons cancelable functions are now implemented
3915 in libc.so as well. No need to call the function in libpthread. This
3916 change allowed to finally disable the incorrect and expensive handling
3917 of weak definition in ld.so.
3918
3919 * Yet more PLT entries in libc.so have been removed. We finally arrived
3920 at the bare minimum. Startup times improved appropriately.
3921
3922 * Support for the new Linux/x86 system call interface was added. The
3923 AT_SYSINFO auxiliary vector entry is recognized and handled.
3924 \f
3925 Version 2.3
3926
3927 * Masahide Washizawa contributed iconv modules for IBM1163 and IBM1164
3928 charsets.
3929
3930 * iconv (the program and the interface) now accepts empty names (excluding
3931 options like //TRANSLIT) to mean "use charset of current locale".
3932
3933 * localedef can now transliterate characters in strings which are not in
3934 the provided charmap. The information from the input locale is used.
3935
3936 * Prelinking support was added for ELF targets. This requires additional
3937 tools and recent versions of the GNU binutils. Contributed by Jakub Jelinek.
3938
3939 * Read-only stdio streams now use mmap to speed up operation by eliminating
3940 copying and buffer underflows. To use add 'm' to the mode string of
3941 the fopen/fdopen/freopen call. Implemented by Ulrich Drepper.
3942
3943 * The malloc functions were completely rewritten by Wolfram Gloger based
3944 on Doug Lea's malloc-2.7.0.c.
3945
3946 * Isamu Hasegawa contributed a completely new and POSIX-conformant
3947 implementation of regex.
3948
3949 * Bruno Haible upgraded the iconv and locale implementation to support
3950 Unicode 3.2.
3951
3952 * Contents of the LC_* and LANG environment variables in the CEN style are
3953 not recognized anymore. It never was used. Change by Ulrich Drepper.
3954
3955 * The runtime (ld.so, libc, libpthread for Linux) now can handle the ELF
3956 thread-local storage (TLS) ABI on some platforms.
3957 Changes by Ulrich Drepper. SH support by Kaz Kojima.
3958
3959 * Bruno Haible contributed iconv converters for ISO-2022-JP-3, SHIFT JIS-X0213,
3960 EUC-JISX0213, and TSCII.
3961
3962 * New header <ifaddrs.h> with functions `getifaddrs' and `freeifaddrs':
3963 BSD-compatible interface for getting all network interface addresses.
3964 Implementation for IPv4 by Roland McGrath.
3965
3966 * Loading of locale data is faster due to the introduction of a locale
3967 archive. Implemented by Roland McGrath and Ulrich Drepper.
3968
3969 * Startup times are significantly reduced by not using exported functions
3970 inside the library itself. Changes by Jakub Jelinek, Roland McGrath,
3971 and Ulrich Drepper.
3972
3973 * Steven Munroe contributed a port to PowerPC64/Linux.
3974 \f
3975 Version 2.2.6
3976
3977 * The Hurd now uses the GNU libio implementation of stdio rather than the
3978 old GNU stdio implementation, and uses a new ABI (libc.so.0.3).
3979
3980 * The Hurd on x86 now has the `ioperm' function and <sys/io.h> header file
3981 with the same behavior as the Linux system call of the same name.
3982 \f
3983 Version 2.2.5
3984
3985 * Stephen Moshier implemented log2, log10, powl and cbrtl for the
3986 128-bit long double format.
3987
3988 * Masahide Washizawa contributed iconv modules for IBM1132, IBM1133, IBM1160,
3989 IBM1161, and IBM1162 charsets.
3990
3991 * Andreas Jaeger contributed a port to x86-64/Linux.
3992
3993 * Peter Bruin contributed a port to PowerPC/Hurd.
3994
3995 * libc functions using I/O streams now can handle wide-oriented streams
3996 as well.
3997
3998 * optimizations in the dynamic linker. Binaries created by recent binutils
3999 versions start up quicker due to reduced time spent on relocations.
4000
4001 * Support for use of gcc3 added by Jakub Jelinek and HJ Lu.
4002 \f
4003 Version 2.2.4
4004
4005 * Stephen Moshier implemented cosh, expm1, log1p, acos, sinh, tanh,
4006 asinh, atanh, j0 for the 128-bit long double format.
4007
4008 * Bruno Haible updated all the code handling Unicode in some form to
4009 support Unicode 3.1.
4010
4011 * Speed of regex for single-byte locales is back to previous levels.
4012 Patch by Isamu Hasegawa.
4013
4014 * Alpha, SPARC, and IA-64 now also using floating stacks.
4015
4016 * Startup time of internationalized applications greatly improved through
4017 iconv cache. Use iconvconfig to generate the cache file.
4018 Contributed by Ulrich Drepper.
4019
4020 * The IA-64 specific part of ld.so was rewritten to eliminate some pretty
4021 severe performance problems. Patch by David Mosberger.
4022
4023 * The Hurd port got a lot more functionality like AIO, various stdio
4024 extensions, etc. Mainly done by Roland McGrath.
4025
4026 * mtrace can now lookup symbols in shared libraries.
4027 \f
4028 Version 2.2.3
4029
4030 * Intel's IA-64 math library is largely integrated. It provides fast and
4031 accurate implementatations for most basic and standard math functions
4032 in float, double, and long double format.
4033
4034 * Stephen Moshier implemented j0, j1, jn, y0, y1, yn, lgamma, erf, erfc,
4035 and asin for the 96-bit long double format and asin, log, tan for the
4036 128-bit long double format.
4037
4038 * The beginning of a last-bit accurate math library by IBM Haifa were added.
4039 The basic double functions exist today. Contributed by Abraham Ziv
4040 <ziv@il.ibm.com>, Moshe Olshansky <olshansk@il.ibm.com>, Ealan Henis
4041 <ealan@il.ibm.com>, and Anna Reitman <reitman@il.ibm.com>.
4042
4043 * An asynchronous name lookup library was added. The interface is designed
4044 after POSIX AIO. The proposal was circulated beforehand to get comments.
4045 No negative ones came in. Implemented by Ulrich Drepper.
4046
4047 * Port to S390/64bit contributed by Martin Schwidefsky
4048 <schwidefsky@de.ibm.com>.
4049
4050 * David Mosberger <davidm@hpl.hp.com> implemented the setcontext family
4051 of functions for Linux/IA-64.
4052
4053 * The RPC code is now thread safe. Threads can now use the same service
4054 of different services at the same time. Patch by Eric Norum
4055 <eric.norum@usask.ca> with some help by Ulrich Drepper.
4056
4057 * Martin Schwidefsky <schwidefsky@de.ibm.com> implemented the setcontext
4058 family of functions for Linux/S390.
4059
4060 * Ulrich Drepper <drepper@redhat.com> implemented the setcontext family
4061 of functions for Linux/x86.
4062
4063 * Port to Linux/CRIS contributed by Axis Communications.
4064 \f
4065 Version 2.2.2
4066
4067 * Lots of headers were cleaned up. Using the tool in the conform/ subdir
4068 we can now check for namespace violations and missing declarations. The
4069 result is that almost all headers are now Unix-compliant (as defined in
4070 the upcoming XPG6). The negative side is that some programs might need
4071 corrections, too, if they depend on the incorrect form of the headers in
4072 previous versions which defined too many symbols and included too many
4073 other headers.
4074
4075 * regex now handles multibyte character sets correctly.
4076 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
4077
4078 * iconv (the program) does now conform to the upcoming XPG6 and handles
4079 charmaps. Instead of the charset names the path of charmaps can be
4080 provided and the conversion happens based on this data.
4081 Contributed by Ulrich Drepper.
4082
4083 * The locale program now provides detailed information about the installed
4084 locales. While
4085
4086 locale -a
4087
4088 only lists the names of the supported locales
4089
4090 locale -a --verbose
4091
4092 provides details such as country, language, and codeset name.
4093 Contributed by Ulrich Drepper.
4094 \f
4095 Version 2.2.1
4096
4097 * The gencat program now parses the input file according to the charset
4098 selected by the LC_CTYPE category. This is important for stateful
4099 character sets. To make generating catalogs easier there is a way
4100 to overwrite the charset selected by the locale: before the first
4101 message or $ quote line the catalog can contain a line like
4102
4103 $ codeset=ISO-8859-2
4104
4105 to select the charset (ISO-8859-2 in this case).
4106
4107 Implemented by Shinya Hanataka and Ulrich Drepper.
4108
4109 * New codeset conversion modules: IBM-922 (Estonia PC codeset),
4110 IBM-1124 (Ukraine PC codeset), IBM-1129 (Vietnamese PC codeset).
4111 Contributed by Masahide Washizawa <washi@jp.ibm.com>.
4112
4113 * Optimized string functions for Alpha ev6 and ev67 by Richard Henderson
4114 <rth@redhat.com> and Rick Gorton <rick.gorton@alpha-processor.com>.
4115
4116 * The LANGUAGE environment variable is now ignored unless the locale is
4117 changed from the default "C" locale.
4118
4119 * The usual bug fixes.
4120 \f
4121 Version 2.2
4122
4123 * Greg McGary added runtime support for bounds checking using gcc's
4124 new -fbounded-pointers option. ix86 target is complete. PowerPC
4125 is in progress.
4126
4127 * Thorsten Kukuk added secure mode support to nscd.
4128
4129 * The Berkeley db libraries have been removed.
4130
4131 Related, the nss_db module is now in a separate package since it
4132 obviously requires a database library being available.
4133
4134 * Wide character I/O streams implemented by Ulrich Drepper.
4135
4136 * Functions from the extended socket API added by Ulrich Drepper.
4137
4138 * Functions feenableexcept and fedisableexcept to control the
4139 behaviour of individual exceptions have been added by Andreas Jaeger.
4140
4141 * ldconfig program added by Andreas Jaeger and Jakub Jelinek.
4142
4143 * The resolver code has been updated from bind 8.2.3-T5B which supports
4144 threads. The integration was done by Andreas Jaeger, Adam D. Bradley,
4145 and Mark Kettenis.
4146
4147 This change could in some situations effect backward compatibility. Since
4148 now `_res' is a thread-local instead of a global variable, modifying it
4149 in one thread does not have any effect in other threads.
4150
4151 The resolver library was also extended to allow IPv6 as the transport
4152 protocol for the requests. This work was done by Stig Venaas.
4153
4154 * Compatibility code for K&R C compilers has been removed from the
4155 header files. A ISO C compiler is needed to use the library
4156 (conforming to either C89 or C99 standard).
4157
4158 * Complete rewrite of the localedef program to support multibyte character
4159 sets. Implement handling of ISO 14651 and ISO 14652. Rewrite strcoll,
4160 strxfrm, wcscoll, and wcsxfrm functions. Make isw*() functions work.
4161 Implemented by Ulrich Drepper.
4162
4163 Bruno Haible significantly improved the generation and use of the data
4164 structures for the wide character tables.
4165
4166 * Plural handling in gettext implemented by Ulrich Drepper.
4167
4168 * The utmp daemon has been removed.
4169
4170 * The port to MIPS-Linux has been finished by Andreas Jaeger.
4171
4172 * A port to Hitachi SH3 and SH4 has been contributed by Kazumoto Kojima
4173 and Yutaka Niibe.
4174
4175 * POSIX clocks and timers implemented by Kaz Kylheku and Ulrich Drepper.
4176
4177 * POSIX barriers implemented by Kaz Kylheku.
4178
4179 * POSIX spawn function family implemented by Ulrich Drepper.
4180
4181 * POSIX spinlocks are now available.
4182
4183 * Timed wait functions for mutex, rwlock, and semaphores are implemented.
4184
4185 * the configure option --enable-kernel=X.Y.Z allows to strip out
4186 compatibility for kernel versions before X.Y.Z. This is currently only
4187 implemented for Linux.
4188
4189 * the sockaddr_in6 structure changed. The IPv6 working group added a new
4190 field sin6_scope_id. This means that all programs using IPv6 should be
4191 recompiled. Don't expect binary compatibility with previous glibc
4192 versions.
4193
4194 * various conversion modules for IBM character sets contributed by
4195 Masahide Washizawa.
4196
4197 * IA-64 port by Jes Sorensen and HJ Lu.
4198 \f
4199 Version 2.1.3
4200
4201 * bug fixes
4202
4203 \f
4204 Version 2.1.2
4205
4206 * bug fixes
4207
4208 \f
4209 Version 2.1.1
4210
4211 * New ISO C 9x function _Exit, imaxabs, and imaxdiv are added.
4212
4213 * New xdr functions are added; some rpc functions are now 64bit clean.
4214
4215 * Fixed a number of bugs and memory leaks (especially in NIS+ code).
4216
4217 * Fixed known incompatibilities with glibc 2.0.
4218
4219 * New functions lock64, strchrnul, rawmemchr, getutmp and getutmpx.
4220
4221 * Optimized a number of functions (especially the ELF dynamic loader).
4222
4223 * Update timezone data files.
4224
4225 * lots of charmaps corrections
4226
4227 * some new locale definitions and charmaps
4228
4229 \f
4230 Version 2.1
4231
4232 * Richard Henderson corrected size of struct timeval on Linux/Alpha to
4233 conform to POSIX member type requirements. Symbol versions have been
4234 adjusted as needed within the library, and for direct use by applications,
4235 but there is potential for problems if third-party libraries use
4236 struct timeval as part of their interface. This does not present
4237 a problem for X and other "essential" system libraries.
4238
4239 * An additional locale model to support C++ Standard Library locale
4240 model and probably more was implemented by Ulrich Drepper.
4241
4242 * Eric Youngdale and Ulrich Drepper implemented versioning of objects on
4243 symbol level.
4244
4245 * Miles Bader provided the `argp' function family to support hierachical
4246 command line argument parsing, layered on top of getopt.
4247
4248 * strtod accepts new hexadecimal floating-point format from ISO C 9X.
4249
4250 * printf has two new specifiers %a and %A to print hexadecimal flaoting-point
4251 numbers.
4252
4253 * scanf recognizes the %a and %A format for scanning floating point numbers.
4254
4255 * the new headers <stdint.h> and <inttypes.h> from ISO C 9X provides
4256 information and interfaces for the available integer types.
4257
4258 * about 130 new math functions were added to implement the ISO C9x math
4259 library.
4260
4261 * the new header <complex.h> contains definitions of the complex math
4262 functions from ISO C 9X.
4263
4264 * the new header <tgmath.h> defines generic macros to use complex or
4265 real valued functions.
4266
4267 * Thorsten Kukuk provided an implementation for NIS+, securelevel 0, 1 and 2.
4268
4269 * Andreas Jaeger provided a test suite for the math library.
4270
4271 * Mark Kettenis implemented the utmpx interface and an utmp daemon.
4272
4273 * Ulrich Drepper added character set conversion functions (iconv).
4274
4275 * Optimized string functions have been added.
4276
4277 * The localedata addon is now part of glibc.
4278
4279 * An implementation of profiling shared libraries was added by Ulrich Drepper.
4280
4281 * Thorsten Kukuk and Ulrich Drepper provided an implementation for a caching
4282 daemon for NSS (nscd).
4283
4284 Missing a better place here are some numbers on improvements. Under
4285 Linux 2.1.125 un-tar-ing the kernel sources takes
4286
4287 user system wall
4288
4289 using local files 12.19s 6.88s 22.91s
4290
4291 using NIS 13.92s 8.91s 26.34s
4292
4293 using NIS & nscd 10.37s 7.34s 25.30s
4294
4295 using NIS+ 27.57s 30.37s 640.46s
4296
4297 using NIS+ & nscd 10.25s 7.83s 26.51s
4298
4299 using NIS & old nscd [1] 13.83s 8.32s 29.60s
4300
4301 Keep in mind that non-namelookup related operations dominate above times.
4302 It was just a common complain that using NIS+ unpacking the kernel is
4303 horribly slow.
4304
4305 [1] The old nscd implementation is not available anymore since it was
4306 distributed with glibc up to version 2.0.98 and thus is now replaced.
4307
4308 * Tim Waugh provided an implementation of the POSIX.2 wordexp function family.
4309
4310 * Mark Kettenis provided a Hesiod NSS module.
4311
4312 * The ELF dynamic loader knows how to process dynamic string tokens ($ORIGIN
4313 and $PLATFORM) in RPATHs and similar strings (Ulrich Drepper).
4314
4315 * rcmd can now handle netgroups (Dick Streefland).
4316
4317 * A port to the ARM was contributed by Phil Blundell, Pat Beirne and Scott
4318 Bambrough.
4319
4320 * Support for the IPv6 protocol has been added to the socket API, as per the
4321 latest draft standards.
4322
4323 * Support for Linux 2.2 has been added.
4324
4325 * Interface changes relative to the latest 2.0.x release:
4326 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4327 addseverity NEW: Unix98
4328 alphasort64 NEW: LFS
4329 argp_err_exit_status NEW: argp, GNU ext
4330 argp_error NEW: argp, GNU ext
4331 argp_failure NEW: argp, GNU ext
4332 argp_help NEW: argp, GNU ext
4333 argp_parse NEW: argp, GNU ext
4334 argp_program_bug_address NEW: argp, GNU ext
4335 argp_program_version NEW: argp, GNU ext
4336 argp_program_version_hook NEW: argp, GNU ext
4337 argp_state_help NEW: argp, GNU ext
4338 argp_usage NEW: argp, GNU ext
4339 authdes_create NEW: Secure RPC
4340 authdes_getucred NEW: Secure RPC
4341 authdes_pk_create NEW: Secure RPC
4342 backtrace NEW: GNU ext.
4343 backtrace_symbols NEW: GNU ext.
4344 backtrace_symbols_fd NEW: GNU ext.
4345 cacos NEW: ISO C 9x
4346 cacosf NEW: ISO C 9x
4347 cacosh NEW: ISO C 9x
4348 cacoshf NEW: ISO C 9x
4349 cacoshl NEW: ISO C 9x
4350 cacosl NEW: ISO C 9x
4351 capget NEW: kernel
4352 capset NEW: kernel
4353 carg NEW: ISO C 9x
4354 cargf NEW: ISO C 9x
4355 cargl NEW: ISO C 9x
4356 casin NEW: ISO C 9x
4357 casinf NEW: ISO C 9x
4358 casinh NEW: ISO C 9x
4359 casinhf NEW: ISO C 9x
4360 casinhl NEW: ISO C 9x
4361 casinl NEW: ISO C 9x
4362 catan NEW: ISO C 9x
4363 catanf NEW: ISO C 9x
4364 catanh NEW: ISO C 9x
4365 catanhf NEW: ISO C 9x
4366 catanhl NEW: ISO C 9x
4367 catanl NEW: ISO C 9x
4368 cbc_crypt NEW: Secure RPC
4369 ccos NEW: ISO C 9x
4370 ccosf NEW: ISO C 9x
4371 ccosh NEW: ISO C 9x
4372 ccoshf NEW: ISO C 9x
4373 ccoshl NEW: ISO C 9x
4374 ccosl NEW: ISO C 9x
4375 cexp NEW: ISO C 9x
4376 cexpf NEW: ISO C 9x
4377 cexpl NEW: ISO C 9x
4378 cimag NEW: ISO C 9x
4379 cimagf NEW: ISO C 9x
4380 cimagl NEW: ISO C 9x
4381 clearerr_locked REMOVED
4382 clntunix_create NEW: sunrpc ext
4383 clog NEW: ISO C 9x
4384 clog10 NEW: ISO C 9x
4385 clog10f NEW: ISO C 9x
4386 clog10l NEW: ISO C 9x
4387 clogf NEW: ISO C 9x
4388 clogl NEW: ISO C 9x
4389 conj NEW: ISO C 9x
4390 conjf NEW: ISO C 9x
4391 conjl NEW: ISO C 9x
4392 cpow NEW: ISO C 9x
4393 cpowf NEW: ISO C 9x
4394 cpowl NEW: ISO C 9x
4395 cproj NEW: ISO C 9x
4396 cprojf NEW: ISO C 9x
4397 cprojl NEW: ISO C 9x
4398 creal NEW: ISO C 9x
4399 crealf NEW: ISO C 9x
4400 creall NEW: ISO C 9x
4401 creat64 NEW: LFS
4402 csin NEW: ISO C 9x
4403 csinf NEW: ISO C 9x
4404 csinh NEW: ISO C 9x
4405 csinhf NEW: ISO C 9x
4406 csinhl NEW: ISO C 9x
4407 csinl NEW: ISO C 9x
4408 csqrt NEW: ISO C 9x
4409 csqrtf NEW: ISO C 9x
4410 csqrtl NEW: ISO C 9x
4411 ctan NEW: ISO C 9x
4412 ctanf NEW: ISO C 9x
4413 ctanh NEW: ISO C 9x
4414 ctanhf NEW: ISO C 9x
4415 ctanhl NEW: ISO C 9x
4416 ctanl NEW: ISO C 9x
4417 des_setparity NEW: Secure RPC
4418 ecb_crypt NEW: Secure RPC
4419 endutxent NEW: Unix98
4420 exp10 NEW: ISO C 9x
4421 exp10f NEW: ISO C 9x
4422 exp10l NEW: ISO C 9x
4423 exp2 NEW: ISO C 9x
4424 exp2f NEW: ISO C 9x
4425 exp2l NEW: ISO C 9x
4426 fattach NEW: STREAMS
4427 fdetach NEW: STREAMS
4428 fdim NEW: ISO C 9x
4429 fdimf NEW: ISO C 9x
4430 fdiml NEW: ISO C 9x
4431 feclearexcept NEW: ISO C 9x
4432 fegetenv NEW: ISO C 9x
4433 fegetexceptflag NEW: ISO C 9x
4434 fegetround NEW: ISO C 9x
4435 feholdexcept NEW: ISO C 9x
4436 feof_locked REMOVED
4437 feraiseexcept NEW: ISO C 9x
4438 ferror_locked REMOVED
4439 fesetenv NEW: ISO C 9x
4440 fesetexceptflag NEW: ISO C 9x
4441 fesetround NEW: ISO C 9x
4442 fetestexcept NEW: ISO C 9x
4443 feupdateenv NEW: ISO C 9x
4444 fflush_locked REMOVED
4445 ffsl NEW: GNU ext.
4446 ffsll NEW: GNU ext.
4447 fgetpos64 NEW: LFS
4448 fgets_unlocked NEW: GNU ext.
4449 fileno_locked REMOVED
4450 fma NEW: ISO C 9x
4451 fmaf NEW: ISO C 9x
4452 fmal NEW: ISO C 9x
4453 fmax NEW: ISO C 9x
4454 fmaxf NEW: ISO C 9x
4455 fmaxl NEW: ISO C 9x
4456 fmin NEW: ISO C 9x
4457 fminf NEW: ISO C 9x
4458 fminl NEW: ISO C 9x
4459 fmtmsg NEW: Unix98
4460 fopen64 NEW: LFS
4461 fputc_locked REMOVED
4462 fputs_unlocked NEW: GNU ext.
4463 fread_unlocked NEW: GNU ext.
4464 freopen64 NEW: LFS
4465 fseeko NEW: Unix98
4466 fsetpos64 NEW: LFS
4467 fstatfs64 NEW: LFS
4468 fstatvfs NEW: Unix98
4469 fstatvfs64 NEW: LFS
4470 ftello NEW: Unix98
4471 ftello64 NEW: LFS
4472 ftruncate64 NEW: LFS
4473 ftw64 NEW: LFS
4474 fwrite_unlocked NEW: GNU ext.
4475 gai_strerror NEW: IPv6
4476 gamma_r REMOVED
4477 gammaf_r REMOVED
4478 gammal_r REMOVED
4479 getchar_locked REMOVED
4480 getdate NEW: Unix98
4481 getdate_err NEW: Unix98
4482 getdate_r NEW: GNU ext.
4483 getmsg NEW: STREAMS
4484 getnameinfo NEW: IPv6
4485 getnetname NEW: Secure RPC
4486 getpmsg NEW: STREAMS
4487 getpt NEW: Unix98 PTY
4488 getrlimit64 NEW: LFS
4489 getutxent NEW: Unix98
4490 getutxid NEW: Unix98
4491 getutxline NEW: Unix98
4492 glob64 NEW: GNU ext.
4493 globfree64 NEW: GNU ext.
4494 gnu_get_libc_release NEW: GNU ext.
4495 gnu_get_libc_version NEW: GNU ext.
4496 grantpt NEW: Unix98 PTY
4497 host2netname NEW: Secure RPC
4498 iconv NEW: iconv
4499 iconv_close NEW: iconv
4500 iconv_open NEW: iconv
4501 if_freenameindex NEW: IPv6
4502 if_indextoname NEW: IPv6
4503 if_nameindex NEW: IPv6
4504 if_nametoindex NEW: IPv6
4505 in6addr_any NEW: IPv6
4506 in6addr_loopback NEW: IPv6
4507 inet6_isipv4mapped NEW: IPv6
4508 isastream NEW: STREAMS
4509 iswblank NEW: GNU ext.
4510 key_decryptsession NEW: Secure RPC
4511 key_decryptsession_pk NEW: Secure RPC
4512 key_encryptsession NEW: Secure RPC
4513 key_encryptsession_pk NEW: Secure RPC
4514 key_gendes NEW: Secure RPC
4515 key_get_conv NEW: Secure RPC
4516 key_secretkey_is_set NEW: Secure RPC
4517 key_setnet NEW: Secure RPC
4518 key_setsecret NEW: Secure RPC
4519 llrint NEW: ISO C 9x
4520 llrintf NEW: ISO C 9x
4521 llrintl NEW: ISO C 9x
4522 llround NEW: ISO C 9x
4523 llroundf NEW: ISO C 9x
4524 llroundl NEW: ISO C 9x
4525 log2 NEW: ISO C 9x
4526 log2f NEW: ISO C 9x
4527 log2l NEW: ISO C 9x
4528 lrint NEW: ISO C 9x
4529 lrintf NEW: ISO C 9x
4530 lrintl NEW: ISO C 9x
4531 lround NEW: ISO C 9x
4532 lroundf NEW: ISO C 9x
4533 lroundl NEW: ISO C 9x
4534 lseek64 NEW: LFS
4535 makecontext NEW: Unix98
4536 mempcpy NEW: GNU ext.
4537 mmap64 NEW: LFS
4538 moncontrol REMOVED
4539 modify_ldt NEW: kernel
4540 nan NEW: ISO C 9x
4541 nanf NEW: ISO C 9x
4542 nanl NEW: ISO C 9x
4543 nearbyint NEW: ISO C 9x
4544 nearbyintf NEW: ISO C 9x
4545 nearbyintl NEW: ISO C 9x
4546 netname2host NEW: Secure RPC
4547 netname2user NEW: Secure RPC
4548 nexttoward NEW: ISO C 9x
4549 nexttowardf NEW: ISO C 9x
4550 nexttowardl NEW: ISO C 9x
4551 nftw NEW: Unix98
4552 nftw64 NEW: LFS
4553 open64 NEW: LFS
4554 passwd2des NEW: Secure RPC
4555 pow10 NEW: GNU ext.
4556 pow10f NEW: GNU ext.
4557 pow10l NEW: GNU ext.
4558 pread NEW: Unix98
4559 pread64 NEW: LFS
4560 printf_size NEW: GNU ext.
4561 printf_size_info NEW: GNU ext.
4562 profil_counter REMOVED
4563 pthread_mutexattr_getkind_np REPLACED
4564 pthread_mutexattr_setkind_np REPLACED
4565 ptsname NEW: Unix98 PTY
4566 ptsname_r NEW: Unix98 PTY
4567 putc_locked REMOVED
4568 putchar_locked REMOVED
4569 putgrent NEW: GNU ext.
4570 putmsg NEW: STREAMS
4571 putpmsg NEW: STREAMS
4572 pututxline NEW: Unix98
4573 pwrite NEW: Unix98
4574 pwrite64 NEW: LFS
4575 readdir64 NEW: LFS
4576 readdir64_r NEW: LFS
4577 remquo NEW: ISO C 9x
4578 remquof NEW: ISO C 9x
4579 remquol NEW: ISO C 9x
4580 round NEW: ISO C 9x
4581 roundf NEW: ISO C 9x
4582 roundl NEW: ISO C 9x
4583 rtime NEW: GNU ext.
4584 scalbln NEW: ISO C 9x
4585 scalblnf NEW: ISO C 9x
4586 scalblnl NEW: ISO C 9x
4587 scandir64 NEW: LFS
4588 sendfile NEW: kernel
4589 setcontext NEW: Unix98
4590 setrlimit64 NEW: LFS
4591 setutxent NEW: Unix98
4592 sighold NEW: Unix98
4593 sigignore NEW: Unix98
4594 sigqueue NEW: POSIX.1b
4595 sigrelse NEW: Unix98
4596 sigset NEW: POSIX.1b
4597 sigtimedwait NEW: POSIX.1b
4598 sigwaitinfo NEW: POSIX.1b
4599 sincos NEW: GNU ext.
4600 sincosf NEW: GNU ext.
4601 sincosl NEW: GNU ext.
4602 statfs64 NEW: LFS
4603 statvfs NEW: Unix98
4604 statvfs64 NEW: LFS
4605 strcasestr NEW: GNU ext.
4606 strtoimax NEW: ISO C 9x
4607 strtoumax NEW: ISO C 9x
4608 strverscmp NEW: GNU ext.
4609 svcauthdes_stats NEW: Secure RPC
4610 svcunix_create NEW: sunrpc etx
4611 svcunixfd_create NEW: sunrpc ext
4612 swapcontext NEW: Unix98
4613 tcgetsid NEW: Unix98 PTY
4614 tdestroy NEW: GNU ext.
4615 tgamma NEW: ISO C 9x
4616 tgammaf NEW: ISO C 9x
4617 tgammal NEW: ISO C 9x
4618 tmpfile64 NEW: LFS
4619 trunc NEW: ISO C 9x
4620 truncate64 NEW: LFS
4621 truncf NEW: ISO C 9x
4622 truncl NEW: ISO C 9x
4623 umount2 NEW: kernel
4624 unlockpt NEW: Unix98 PTY
4625 updwtmpx NEW: Unix98
4626 user2netname NEW: Secure RPC
4627 utmpxname NEW: Unix98
4628 versionsort NEW: GNU ext.
4629 versionsort64 NEW: GNU ext.
4630 waitid NEW: Unix98
4631 wcscasecmp NEW: GNU ext.
4632 wcsncasecmp NEW: GNU ext.
4633 wcsnlen NEW: GNU ext.
4634 wcstoimax NEW: ISO C 9x
4635 wcstoll NEW: ISO C 9x
4636 wcstoull NEW: ISO C 9x
4637 wcstoumax NEW: ISO C 9x
4638 wcswcs NEW: Unix98
4639 wordexp NEW: POSIX.2
4640 wordfree NEW: POSIX.2
4641 write_profiling REMOVED
4642 xdecrypt NEW: Secure RPC
4643 xdr_authdes_cred NEW: Secure RPC
4644 xdr_authdes_verf NEW: Secure RPC
4645 xdr_cryptkeyarg NEW: Secure RPC
4646 xdr_cryptkeyarg2 NEW: Secure RPC
4647 xdr_cryptkeyres NEW: Secure RPC
4648 xdr_getcredres NEW: Secure RPC
4649 xdr_key_netstarg NEW: Secure RPC
4650 xdr_key_netstres NEW: Secure RPC
4651 xdr_keybuf NEW: Secure RPC
4652 xdr_keystatus NEW: Secure RPC
4653 xdr_netnamestr NEW: Secure RPC
4654 xdr_sizeof NEW: Secure RPC
4655 xdr_unixcred NEW: sunrpc ext
4656 xencrypt NEW: Secure RPC
4657 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
4658 \f
4659 Version 2.0.6
4660
4661 * more bug fixes
4662
4663 \f
4664 Version 2.0.5
4665
4666 * more bug fixes
4667
4668 * inet_ntoa is thread-safe
4669
4670 * updwtmp is moved from libutil to libc
4671
4672 * rewrite of cbrt function
4673
4674 * update of timezone data
4675 \f
4676 Version 2.0.4
4677
4678 * more bug fixes
4679 \f
4680 Version 2.0.3
4681
4682 * more bug fixes
4683 \f
4684 Version 2.0.2
4685
4686 * more bug fixes
4687
4688 * add atoll function
4689
4690 * fix complex problems in Berkeley DB code
4691
4692 * fix math functions
4693 \f
4694 Version 2.0.1
4695
4696 * fixed lots of header problems (especially Linux/GNU specific)
4697
4698 * dynamic loader preserves all registers
4699
4700 * Roland McGrath provided support for handling of auxiliary objects in
4701 the ELF dynamic loader.
4702
4703 * support for parallel builds is improved
4704 \f
4705 Version 2.0
4706
4707 * GNU extensions are no longer declared by default. To enable them you
4708 must define the macro `_GNU_SOURCE' in your program or compile with
4709 `-D_GNU_SOURCE'.
4710
4711 * The library has changed from using GNU ld symbol aliases to using weak
4712 symbols where available. The ELF object file format supports weak
4713 symbols; GNU ld also supports weak symbols in the a.out format. (There
4714 is also now support for other GNU ld extensions in ELF. Use the
4715 `--with-elf' option to configure to indicate you have ELF, and
4716 `--with-gnu-ld' if using GNU ld.) This change resulted in the deletion
4717 of many files which contained only symbol aliases, reducing the size of
4718 the source and the compiled library; many other files were renamed to
4719 less cryptic names previously occupied by the symbol alias files.
4720 There is a new header file <elf.h> for programs which operate on
4721 files in the ELF format.
4722
4723 * Converted to Autoconf version 2, so `configure' has more options.
4724 Run `configure --help' to see the details.
4725
4726 * The library can now be configured to build profiling, highly-optimized
4727 (but undebuggable), and/or shared libraries (ELF with GNU ld only). The
4728 `--enable-profile', `--enable-omitfp', and `--enable-shared' options to
4729 `configure' enable building these extra libraries. The shared library is
4730 built by default when using both ELF and GNU ld. When shared libraries
4731 are enabled, the new library `-ldl' is available for arbitrary run-time
4732 loading of shared objects; its interface is defined in <dlfcn.h>. The
4733 new header file <link.h> gives access to the internals of the run-time
4734 dynamic linker, `ld.so'. The shell script `ldd' is similar to the
4735 application of same name on other systems and it provides information
4736 about dynamically linked binaries.
4737
4738 * The C library now provides the run-time support code for profiling
4739 executables compiled with `-pg'. Programs can control the profiling code
4740 through the interface in <sys/gmon.h>. The `gmon.out' files written by
4741 the GNU C library can be read only by GNU `gprof' (from GNU binutils);
4742 the support for this file format was contributed by David Mosberger-Tang.
4743
4744 * The math code has been replaced with a math library based on fdlibm from
4745 Sun, and modified by JT Conklin and Ulrich Drepper with i387 support, by
4746 Ian Taylor with `float' functions and by Ulrich Drepper with `long double'
4747 functions. The math functions now reside in a separate library, so
4748 programs using them will need to use `-lm' their linking commands.
4749
4750 * John C. Bowman contributed optimized ix87 assembler inline functions.
4751
4752 * Ulrich Drepper has contributed support for an `/etc/nsswitch.conf'
4753 mechanism similar to that found in Solaris 2. This is now used for the
4754 group, passwd, hosts, networks, services, protocols, rpc, ethers,
4755 shadow, netgroup, publickey, and alias databases. The `nsswitch.conf'
4756 file controls what services are used for each individual database. This
4757 works by loading shared libraries with names specified in `nsswitch.conf',
4758 so service modules can be changed or added at any time without even
4759 relinking any program. Currently there are the file, db, and NIS based
4760 NSS services available.
4761
4762 * The new functions `strtoq' and `strtouq' parse integer values from
4763 strings, like `strtol' and `strtoul', but they return `long long int' and
4764 `unsigned long long int' values, respectively (64-bit quantities).
4765
4766 * The new functions `strtof' and `strtold' parse floating-point values from
4767 strings, like `strtod', but they return `float' and `long double' values,
4768 respectively (on some machines `double' and `long double' are the same).
4769
4770 * Ulrich Drepper has contributed new implementations of the floating-point
4771 printing and reading code used in the `printf' family of functions and
4772 `strtod', `strtof', and `strtold'. These new functions are perfectly
4773 accurate, and much faster than the old ones.
4774
4775 * The implementation of the POSIX locale model was completely rewritten by
4776 Ulrich Drepper. This includes the new programs `localedef' and `locale'
4777 to compile the POSIX locale definition.
4778
4779 * The former dummy implementations of the strcoll and strxfrm function are
4780 now replaced by fully functional code contributed by Ulrich Drepper. The
4781 collation information comes from the POSIX locale definitions.
4782
4783 * The new header <langinfo.h> defines an interface for accessing
4784 various locale-dependent data (using the locale chosen with `setlocale').
4785
4786 * Ulrich Drepper has contributed a new suite of functions for operation on
4787 wide-character and multibyte-character strings, in <wchar.h>;
4788 and classification and case conversion of wide characters, in <wctype.h>.
4789 These new functions are conforming to the ISO C, Amendement 1 specification.
4790
4791 * There is now a second implementation of the standard I/O library available.
4792 It comes from GNU libg++ as was written by Per Bothner, heavily modified
4793 by Hongjiu Lu and made thread safe by Ulrich Drepper.
4794
4795 * You can now use positional parameter specifications in format strings
4796 for the `printf' and `scanf' families of functions. For example,
4797 `printf ("Number %2$d, Mr %1$s\n", "Jones", 6);'' prints
4798 ``Number 6, Mr Jones''. This is mainly useful when providing different
4799 format strings for different languages, whose grammars may dictate
4800 different orderings of the values being printed. To support this
4801 feature, the interface for `register_printf_handler' has changed; see
4802 the header file <printf.h> for details.
4803
4804 * The `printf' and `scanf' families of functions now understand a new
4805 formatting flag for numeric conversions: the ' flag (e.g. %'d or %'f) says
4806 to group numbers as indicated by the locale; for `scanf' and friends, this
4807 says to accept as valid only a number with all the proper grouping
4808 separators in the right places. In the default "C" locale, numbers are
4809 not grouped; but locales for specific countries will define the usual
4810 conventions (i.e. separate thousands with `,' in the US locale).
4811
4812 * The pgrp functions have been regularized, slightly incompatibly but much
4813 less confusingly. The core functions are now `getpgid' and `setpgid',
4814 which take arguments for the PID to operate on; the POSIX.1 `getpgrp' (no
4815 argument) and BSD `setpgrp' (identical to `setpgid') functions are
4816 provided for compatibility. There is no longer an incompatible `getpgrp'
4817 with an argument declared under _BSD_SOURCE; no BSD code uses it.
4818
4819 * The new header file <fts.h> and suite of functions simplify programs that
4820 operate on directory trees. This code comes from 4.4 BSD.
4821
4822 * The resolver code has been updated from the BIND 4.9.5-P1 release.
4823 Parts of the code were heavily modified by Ulrich Drepper to fit in the
4824 NSS scheme used in glibc.
4825
4826 * There is a new malloc debugging hook `__memalign_hook'.
4827
4828 * There are new typedefs `ushort' for `unsigned short int' and `uint' for
4829 `unsigned int' in <sys/types.h>. These are for compatibility only and
4830 their use is discouraged.
4831
4832 * The `-lmcheck' library to enable standard malloc debugging hooks is now
4833 done differently, so that it works even without GNU ld.
4834
4835 * New function `euidaccess' checks allowed access to a file like `access',
4836 but using the effective IDs instead of the real IDs.
4837
4838 * The time zone data files have been updated for the latest and greatest
4839 local time conventions of the countries of the world.
4840
4841 * The new function `dirfd' extracts the file descriptor used by a DIR stream;
4842 see <dirent.h>.
4843
4844 * The new functions `ecvt', `fcvt', and `gcvt' provide an obsolete interface
4845 for formatting floating-point numbers. They are provided only for
4846 compatibility; new programs should use `sprintf' instead. There are
4847 also equivalent function for the `long double' floating-point type and
4848 all functions also exist in a reentrant form.
4849
4850 * The new auxiliary library `-lutil' from 4.4 BSD contains various
4851 functions for maintaining the login-record files (primarily of use to
4852 system programs such as `login'), and convenient functions for
4853 allocating and initializing a pseudo-terminal (pty) device.
4854
4855 * Ulrich Drepper has contributed new support for System V style
4856 shared memory and IPC on systems that support it.
4857
4858 * Ulrich Drepper has contributed several miscellaneous new functions found
4859 in System V: The `hsearch' family of functions provide an effective
4860 implementation of hash tables; `a64l' and `l64a' provide a very simple
4861 binary to ASCII mapping; `drand48' and friends provide a 48-bit random
4862 number generator.
4863
4864 * Ulrich Drepper has contributed new reentrant counterparts for the
4865 `random' and `hsearch' families of functions; `random_r', `hsearch_r', etc.
4866
4867 * Ulrich Drepper has contributed new, highly-optimized versions of several
4868 string functions for the i486/Pentium family of processors.
4869
4870 * Ulrich Drepper has updated the Linux-specific code, based largely
4871 on work done in Hongjiu Lu's version of GNU libc for Linux.
4872 The GNU library now supports Linux versions 2.0.10 and later,
4873 using the ELF object file format (i[3456]86-*-linux).
4874
4875 * Andreas Schwab has ported the C library to Linux/m68k (m68k-*-linux).
4876
4877 * David Mosberger-Tang and Richard Henderson have ported the C library
4878 to Linux/Alpha (alpha-*-linux). Richard Henderson contributed the
4879 dynamic linking support for ELF/Alpha.
4880
4881 * Richard Henderson contributed several Alpha optimized assembler function
4882 for arithmetic and string handling.
4883
4884 * Ulrich Drepper has contributed a new set of message catalog functions to
4885 support multiple languages using the <libintl.h> interface, for use with
4886 his new package GNU gettext. Translation volunteers have contributed
4887 catalogs of the library's messages in Spanish, German, and Korean.
4888
4889 * For compatibility with XPG4, Ulrich Drepper has contributed the `gencat'
4890 program and the `catgets' function for reading the catalog files it
4891 creates. (The <libintl.h> interface is preferred; we include the
4892 <nl_types.h> interface using `catgets' only for source compatibility with
4893 programs already written to use it.)
4894
4895 * New header file <values.h> gives SVID-compatible names for <limits.h>
4896 constants.
4897
4898 * Various new macros, declarations, and small header files for compatibility
4899 with 4.4 BSD.
4900
4901 * New function `group_member' is a convenient way to check if a process has
4902 a given effective group ID.
4903
4904 * When using GCC 2.7 and later, the socket functions are now declared in a
4905 special way so that passing an argument of type `struct sockaddr_in *',
4906 `struct sockaddr_ns *', or `struct sockaddr_un *' instead of the generic
4907 `struct sockaddr *' type, does not generate a type-clash warning.
4908
4909 * New function `error' declared in header file <error.h> is a convenient
4910 function for printing error messages and optionally exiting; this is the
4911 canonical function used in GNU programs. The new functions `err', `warn',
4912 and friends in header file <err.h> are the canonical 4.4 BSD interface for
4913 doing the same thing.
4914
4915 * The <glob.h> interface has several new flags from 4.4 BSD that extend the
4916 POSIX.2 `glob' function to do ~ and {...} expansion.
4917
4918 * New function `unsetenv' complements `setenv' for compatibility with 4.4 BSD.
4919 `clearenv' which is used in POSIX.9 is also available.
4920
4921 * New function `getsid' returns session ID number on systems that support it.
4922
4923 * We have incorporated the 4.4 BSD `db' library (version 1.85). New header
4924 files <db.h> and <mpool.h> provide a rich set of functions for several
4925 types of simple databases stored in memory and in files, and <ndbm.h> is
4926 an old `ndbm'-compatible interface using the `db' functions. Link with
4927 `-ldb' to get these functions.
4928
4929 * New macro `strdupa' copies a string like `strdup', but uses local stack
4930 space from `alloca' instead of dynamic heap space from `malloc'.
4931
4932 * New function `strnlen' is like `strlen' but searches only a given maximum
4933 number of characters for the null terminator. `stpncpy', `strndup' and
4934 `strndupa' are similar variants for the `stpcpy', `strdup' and `strdupa'
4935 function.
4936
4937 * New function `statfs' in header <sys/statfs.h>.
4938
4939 * The new <argz.h> and <envz.h> interfaces contributed by Miles Bader
4940 provide convenient functions for operating on blocks of null-terminated
4941 strings.
4942
4943 * A new suite of functions in <utmp.h> handle all the details of reading
4944 and writing the utmp file.
4945
4946 * An implementation of the NIS/YP(tm) based NSS service was contributed by
4947 Thorsten Kukuk.
4948
4949 * Paul Eggert and Ulrich Drepper modified the `strftime' function to be
4950 completely POSIX compliant and also implemented the extended functionality
4951 to handle alternate digit representation and alternate era date formats.
4952
4953 * Ulrich Drepper provided an implementation of the `strptime' function
4954 defined in XPG4.2 which transforms a string into a `struct tm' value.
4955
4956 * Paul Eggert provided the tzselect shell script as part of the timezone
4957 code. The shell script makes it easy to select the correct timezone
4958 specification.
4959
4960 * The implementation of the malloc family of functions is completely replaced
4961 by a new implementation by Doug Lea with many improvements by Wolfram Gloger.
4962 The implementation uses the mmap function (if available) and it is
4963 optimized for the use in multi threaded programs.
4964
4965 * Ulrich Drepper contributed a MD5 "encryption" for the crypt family of
4966 functions. This new functionality is usable by specifying a special
4967 salt string and it is compatible with implementation on *BSD systems.
4968
4969 * Lots of functions from the XPG4.2 standard were added by Ulrich Drepper:
4970 `getsubopt' to handle second level command line options, `bsd_signal'
4971 to access BSD style `signal' functionality, the obsolete `regexp' style
4972 expression matcher.
4973
4974 * the `lchown' function is available on system which support this
4975 functionality.
4976
4977 * The implementation of the shadow password handling function was contributed
4978 by Ulrich Drepper.
4979
4980 * David Mosberger-Tang changed the SunRPC implementation to be 64bit safe.
4981
4982 * POSIX.1g support was added. The <sys/select.h> header is available,
4983 `isfdtype' and `pselect' are implemented. Craig Metz contributed an
4984 implementation of `getaddrinfo'.
4985 \f
4986 Version 1.09
4987
4988 * For cross-compilation you should now set `BUILD_CC' instead of `HOST_CC'.
4989
4990 * New header file <fstab.h> and new functions `getfsspec', `getfsent' and
4991 friends, for parsing /etc/fstab. This code comes from 4.4 BSD.
4992
4993 * The new function `daemon' from 4.4 BSD is useful for server programs that
4994 want to put themselves in the background.
4995
4996 * Joel Sherrill has contributed support for several standalone boards that
4997 run without an operating system.
4998
4999 * `printf', `scanf' and friends now accept a `q' type modifier for long
5000 long int as well as `ll'. Formats using these might be `%qu' or `%lld'.
5001
5002 * All of the code taken from BSD (notably most of the math and networking
5003 routines) has been updated from the BSD 4.4-Lite release.
5004
5005 * The resolver code has been updated from the BIND-4.9.3-BETA9 release.
5006
5007 * The new functions `getdomainname' and `setdomainname' fetch or change the
5008 YP/NIS domain name. These are system calls which exist on systems which
5009 have YP (aka NIS).
5010
5011 * The time zone data files have been updated for the latest international
5012 conventions.
5013
5014 * The SunRPC programs `portmap' and `rpcinfo' are now installed in
5015 $(sbindir) (usually /usr/local/sbin) instead of $(bindir).
5016 \f
5017 Version 1.08
5018
5019 * The C library now includes support for Sun RPC, from Sun's free
5020 RPCSRC-4.0 distribution. The `portmap', `rpcinfo', and `rpcgen' programs
5021 are included. (There is still no support for YP.)
5022
5023 * Tom Quinn has contributed a port of the C library to SGI machines running
5024 Irix 4 (mips-sgi-irix4).
5025
5026 * The new `lockf' function is a simplified interface to the locking
5027 facilities of `fcntl', included for compatibility.
5028
5029 * New time functions `timegm', `timelocal', and `dysize' for compatibility.
5030
5031 * New header file <sys/timeb.h> and new function `ftime' for compatibility.
5032
5033 * New header files <poll.h> and <sys/poll.h> and new function `poll' for
5034 compatibility.
5035
5036 * The error message printed by `assert' for a failed assertion now includes
5037 the name of the program (if using GNU ld) and the name of the calling
5038 function (with versions of GCC that support this).
5039
5040 * The `psignal' function is now declared in <signal.h>, not <stdio.h>.
5041
5042 * The library now includes the <sys/mman.h> header file and memory
5043 management functions `mmap', `munmap', `mprotect', `msync', and
5044 `madvise', on systems that support those facilities.
5045
5046 * The interface for `mcheck' has changed slightly: the function called to
5047 abort the program when an allocation inconsistency is detected now takes
5048 an argument that indicates the type of failure. The new function
5049 `mprobe' lets you request a consistency check for a particular block at
5050 any time (checks are normally done only when you call `free' or `realloc'
5051 on a block).
5052
5053 * It is now possible to easily cross-compile the C library, building on one
5054 system a library to run on another machine and/or operating system. All
5055 you need to do is set the variable `HOST_CC' in `configparms' to the
5056 native compiler for programs to run on the machine you are building on (a
5057 few generator programs are used on Unix systems); set `CC' to the
5058 cross-compiler.
5059
5060 * The new function `fexecve' (only implemented on the GNU system) executes
5061 a program file given a file descriptor already open on the file.
5062 \f
5063 Version 1.07
5064
5065 * Brendan Kehoe has contributed most of a port to the DEC Alpha
5066 running OSF/1 (alpha-dec-osf1). He says it is 75% complete.
5067
5068 * You can set the variable `libprefix' in `configparms' to specify a prefix
5069 to be prepended to installed library files; this makes it easy to install
5070 the GNU C library to be linked as `-lgnuc' or whatever.
5071
5072 * The new `stpncpy' is a cross between `stpcpy' and `strncpy': It
5073 copies a limited number of characters from a string, and returns the
5074 address of the last character written.
5075
5076 * You no longer need to check for whether the installed `stddef.h' is
5077 compatible with the GNU C library. configure now checks for you.
5078
5079 * You can now define a per-stream `fileno' function to convert the
5080 stream's cookie into an integral file descriptor.
5081
5082 * ``malloc (0)'' no longer returns a null pointer. Instead, it
5083 allocates zero bytes of storage, and returns a unique pointer which
5084 you can pass to `realloc' or `free'. The behavior is undefined if
5085 you dereference this pointer.
5086
5087 * The C library now runs on Sony NEWS m68k machines running either
5088 NewsOS 3 or NewsOS 4.
5089
5090 * The new `syscall' function is a system-dependent primitive function
5091 for invoking system calls. It has the canonical behavior on Unix
5092 systems, including unreliable return values for some calls (such as
5093 `pipe', `fork' and `getppid').
5094
5095 * The error code `EWOULDBLOCK' is now obsolete; it is always defined
5096 to `EAGAIN', which is the preferred name. On systems whose kernels
5097 use two distinct codes, the C library now translates EWOULDBLOCK to
5098 EAGAIN in every system call function.
5099 \f
5100 Version 1.06
5101
5102 * The GNU C Library Reference Manual is now distributed with the library.
5103 `make dvi' will produce a DVI file of the printed manual.
5104 `make info' will produce Info files that you can read on line using C-h i
5105 in Emacs or the `info' program.
5106 Please send comments on the manual to bug-glibc-manual@gnu.org.
5107
5108 * The library now supports SVR4 on i386s (i386-unknown-sysv4).
5109
5110 * Brendan Kehoe has contributed a port to Sun SPARCs running Solaris 2.
5111
5112 * Jason Merrill has contributed a port to the Sequent Symmetry running
5113 Dynix version 3 (i386-sequent-dynix).
5114
5115 * The library has been ported to i386s running SCO 3.2.4 (also known as SCO
5116 ODT 2.0; i386-unknown-sco3.2.4) or SCO 3.2 (i386-unknown-sco3.2).
5117
5118 * New function `memory_warnings' lets you arrange to get warnings when
5119 malloc is running out of memory to allocate, like Emacs gives you.
5120
5121 * The C library now contains the relocating allocator used in Emacs 19 for
5122 its editing buffers. This allocator (ralloc) minimizes allocation
5123 overhead and fragmentation by moving allocated regions around whenever it
5124 needs to. You always refer to a ralloc'd region with a "handle" (a
5125 pointer to a pointer--an object of type `void **').
5126
5127 * There is a new `printf' format: `%m' gives you the string corresponding
5128 to the error code in `errno'.
5129
5130 * In `scanf' formats, you can now use `%as' or `%a[' to do the normal `%s'
5131 or `%[' conversion, but instead of filling in a fixed-sized buffer you
5132 pass, the `a' modifier says to fill in a `char **' you pass with a
5133 malloc'd string.
5134
5135 * The `fnmatch' function supports the new flag bits `FNM_LEADING_DIR' and
5136 `FNM_CASEFOLD'. `FNM_LEADING_DIR' lets a pattern like `foo*' match a
5137 name like `foo/bar'. `FNM_CASEFOLD' says to ignore case in matching.
5138
5139 * `mkstemp' is a traditional Unix function to atomically create and open a
5140 uniquely-named temporary file.
5141 \f
5142 Version 1.05
5143
5144 * The standard location for the file that says what the local timezone is
5145 has changed again. It is now `/usr/local/etc/localtime' (or more
5146 precisely, `${prefix}/etc/localtime') rather than `/etc/localtime'.
5147
5148 * The distribution no longer contains any files with names longer than 14
5149 characters.
5150
5151 * `struct ttyent' has two new flag bits: TTY_TRUSTED and TTY_CONSOLE.
5152 These are set by the new `trusted' and `console' keywords in `/etc/ttys'.
5153
5154 * New functions `ttyslot' and `syslog' from 4.4 BSD.
5155 \f
5156 Version 1.04
5157
5158 * The configuration process has changed quite a bit. The `configure'
5159 script is now used just like the configuration scripts for other GNU
5160 packages. The `sysdeps' directory hierarchy is much rearranged.
5161 The file `INSTALL' explains the new scheme in detail.
5162
5163 * The header files no longer need to be processed into ANSI C and
5164 traditional C versions. There is just one set of files to install, and
5165 it will work with ANSI or old C compilers (including `gcc -traditional').
5166
5167 * Brendan Kehoe and Ian Lance Taylor have ported the library to the
5168 MIPS DECStation running Ultrix 4.
5169
5170 * The Sun 4 startup code (crt0) can now properly load SunOS 4 shared libraries.
5171 Tom Quinn contributed the initial code. The GNU C library can NOT yet be
5172 made itself into a shared library.
5173
5174 * Yet further improved support for the i386, running 4.3 BSD-like systems
5175 (such as Mach 3 with the Unix single-server), or System V.
5176
5177 * New function `strncasecmp' to do case-insensitive string comparison
5178 with limited length.
5179
5180 * New function `strsep' is a reentrant alternative to `strtok'.
5181
5182 * New functions `scandir' and `alphasort' for searching directories.
5183
5184 * New function `setenv' is a better interface to `putenv'.
5185
5186 * Ian Lance Taylor has contributed an implementation of the SVID `ftw'
5187 function for traversing a directory tree.
5188
5189 * The GNU obstack package is now also part of the C library.
5190 The new function `open_obstack_stream' creates a stdio stream that
5191 writes onto an obstack; `obstack_printf' and `obstack_vprintf' do
5192 formatted output directly to an obstack.
5193
5194 * Miscellaneous new functions: reboot, nice, sigaltstack (4.4 BSD only),
5195 cfmakeraw, getusershell, getpass, swab, getttyent, seteuid, setegid.
5196
5197 * `FNM_FILE_NAME' is another name for `FNM_PATHNAME', used with `fnmatch'.
5198
5199 * The new functions `strfry' and `memfrob' do mysterious and wonderful
5200 things to your strings.
5201
5202 * There are some new test programs: test-fseek, testmb, and testrand.
5203
5204 * Some work has been done to begin porting the library to 4.4 BSD and Linux.
5205 These ports are not finished, but are a good starting place for really
5206 supporting those systems.
5207
5208 * `/etc/localtime' is now the standard location for the file that says what
5209 the local timezone is, rather than `/usr/local/lib/zoneinfo/localtime'.
5210 This follows the general principle that `/etc' is the place for all local
5211 configuration files.
5212
5213 * The C library header files now use `extern "C"' when used by the C++
5214 compiler, so the C library should now work with C++ code.
5215
5216 * The header file <bstring.h> is gone. <string.h> now declares bcopy,
5217 bcmp, bzero, and ffs. (Update: nowadays these functions are declared
5218 in <strings.h>.)
5219
5220 * Mike Haertel (of GNU e?grep and malloc fame) has written a new sorting
5221 function which uses the `merge sort' algorithm, and is said to be
5222 significantly faster than the old GNU `qsort' function. Merge sort is
5223 now the standard `qsort' function. The new algorithm can require a lot
5224 of temporary storage; so, the old sorting function is called when the
5225 required storage is not available.
5226
5227 * The C library now includes Michael Glad's Ultra Fast Crypt, which
5228 provides the Unix `crypt' function, plus some other entry points.
5229
5230 * The code and header files taken from 4.4 BSD have been updated with the
5231 latest files released from Berkeley.
5232 \f
5233 ----------------------------------------------------------------------
5234 Copyright information:
5235
5236 Copyright (C) 1992-2018 Free Software Foundation, Inc.
5237
5238 Permission is granted to anyone to make or distribute verbatim copies
5239 of this document as received, in any medium, provided that the
5240 copyright notice and this permission notice are preserved,
5241 thus giving the recipient permission to redistribute in turn.
5242
5243 Permission is granted to distribute modified versions
5244 of this document, or of portions of it,
5245 under the above conditions, provided also that they
5246 carry prominent notices stating who last changed them.
5247 \f
5248 Local variables:
5249 version-control: never
5250 fill-column: 76
5251 End: